Simulation of log fitness change
Laure Olazcuaga, Nicolas Rode
11-janvier-2021
1 DATA
## Longitudinal data
data_sum <- loadlongitudinaldata(dataset = "DATA_Adults_G1G29.csv", rm_generation1 = 1,rm_generation2 = 7,rm_generation3 = 29)
## Phenotyping steps
data_G0 <- loadfitnessdata(dataset = "Selection_Phenotypage_G0_G7_G8.csv", generation = "G1")
data_G7 <- loadfitnessdata(dataset = "Selection_Phenotypage_G0_G7_G8.csv", generation = "G7")
data_G29 <- loadfitnessdata(dataset =
"PERFORMANCE_Comptage_adultes_G13G14G15G16G17G18G19G20G21G22G23G24G25G26G27G28G29.csv",
generation = "29")
head(data_sum)## Line Fruit_s Generation Phase N Nb_adults sd fitness
## 1 CE1 Cherry 2 first_prepool 20 9.15000 6.123939 -0.7819784
## 2 CE1 Cherry 3 first_prepool 10 15.30000 7.631077 -0.2678794
## 3 CE1 Cherry 4 first_prepool 8 15.00000 7.782765 -0.2876821
## 4 CE1 Cherry 5 first_prepool 6 14.50000 6.284903 -0.3215836
## 5 CE2 Cherry 2 first_prepool 20 7.75000 7.758696 -0.9480394
## 6 CE2 Cherry 3 first_prepool 7 14.42857 8.303757 -0.3265219
## se_fitness
## 1 0.1496562
## 2 0.1577228
## 3 0.1834415
## 4 0.1769518
## 5 0.2238577
## 6 0.2175215head(data_G0)## Treatment Line Fruit_s Nb_eggs Nb_adults SA Emergence_rate
## 993 Cherry CE1 GF 76 6 0 0.07894737
## 994 Cherry CE1 GF 89 17 0 0.19101124
## 995 Cherry CE1 GF 57 12 0 0.21052632
## 996 Cherry CE1 GF 172 24 0 0.13953488
## 997 Cherry CE1 GF 173 33 0 0.19075145
## 998 Cherry CE1 GF 91 18 0 0.19780220head(data_G7)## Treatment Line Fruit_s Nb_eggs Nb_adults SA Emergence_rate
## 3 Strawberry CR4 Cranberry 152 68 0 0.4473684
## 4 Cranberry CR4 Cranberry 246 25 1 0.1016260
## 5 Cherry CR4 Cranberry 238 29 0 0.1218487
## 6 Cherry CR4 Cranberry 166 23 0 0.1385542
## 8 Cranberry FR3 Strawberry 204 5 0 0.0245098
## 9 Strawberry FR3 Strawberry 124 45 1 0.3629032head(data_G29)## Treatment Line Fruit_s Nb_eggs Nb_adults SA Emergence_rate
## 5392 Strawberry CEA Cherry 196 16 0 0.08163265
## 5393 Strawberry CEA Cherry 192 30 0 0.15625000
## 5394 Strawberry CEA Cherry 160 17 0 0.10625000
## 5395 Strawberry CEA Cherry 106 9 0 0.08490566
## 5396 Strawberry CEA Cherry 119 14 0 0.11764706
## 5397 Strawberry CEA Cherry 204 24 0 0.11764706dim(data_G29)## [1] 990 7#Compute log change
## !! sd_logchange are standard errors, not standard deviations
data_logchange <- rbind(computelogchangenew(fitness_dataset_ancestral = data_G0, fitness_dataset_evolved = data_G7, trait = "Nb_adults", gen=7), computelogchangenew(fitness_dataset_ancestral = data_G0, fitness_dataset_evolved = data_G29, trait = "Nb_adults", gen=29))
data_logchangeNb_eggs <- rbind(computelogchangenew(fitness_dataset_ancestral = data_G0, fitness_dataset_evolved = data_G7, trait = "Nb_eggs", gen=7), computelogchangenew(fitness_dataset_ancestral = data_G0, fitness_dataset_evolved = data_G29, trait = "Nb_eggs", gen=29))
data_logchangeEmergence_rate <- rbind(computelogchangenew(fitness_dataset_ancestral = data_G0, fitness_dataset_evolved = data_G7, trait = "Emergence_rate", gen=7), computelogchangenew(fitness_dataset_ancestral = data_G0, fitness_dataset_evolved = data_G29, trait = "Emergence_rate", gen=29))
## Combine datasets
data_logchangeNb_eggsEmergence_rate <- data.frame(data_logchange, Nb_eggslogchange=data_logchangeNb_eggs$logchange, sdNb_eggslogchange=data_logchangeNb_eggs$sd_logchange, Emergence_ratelogchange=data_logchangeEmergence_rate$logchange, sdEmergence_ratelogchange=data_logchangeEmergence_rate$sd_logchange)
#Lines: CE2 and CR2 do not pass Geary's test (the threshold of a standardized mean greater than 3)
data_logchange$Line_type <- ifelse(data_logchange$Line == "CR2"| data_logchange$Line == "CE2", "solid","dashed")
data_logchangeNb_eggs$Line_type <- ifelse(data_logchange$Line == "CR2"| data_logchange$Line == "CE2", "solid","dashed")
data_logchangeEmergence_rate$Line_type <- ifelse(data_logchange$Line == "CR2"| data_logchange$Line == "CE2", "solid","dashed")
#Formatting for testing correlation
TEMP_dataG7_CheCran <- formattinglogchange(data_logchange, "7", "Cherry", "Cranberry")
TEMP_dataG7_CranStraw <- formattinglogchange(data_logchange, "7", "Cranberry", "Strawberry")
TEMP_dataG7_StrawChe <- formattinglogchange(data_logchange, "7", "Strawberry", "Cherry")
TEMP_dataG29_CheCran <- formattinglogchange(data_logchange, "29", "Cherry", "Cranberry")
TEMP_dataG29_CranStraw <- formattinglogchange(data_logchange, "29", "Cranberry", "Strawberry")
TEMP_dataG29_StrawChe <- formattinglogchange(data_logchange, "29", "Strawberry", "Cherry")2 ADAPTATION
2.1 Proportion of adaptation: raw data
tapply(data_sum$Nb_adults[data_sum$Generation=="2"],data_sum$Fruit_s[data_sum$Generation=="2"],mean)## Blackcurrant Cherry Cranberry Fig Grape Rosehips
## NA 8.80000 19.38667 NA NA NA
## Strawberry Tomato
## 12.19000 NAtapply(data_sum$Nb_adults[data_sum$Generation=="27"],data_sum$Fruit_s[data_sum$Generation=="27"],mean)## Blackcurrant Cherry Cranberry Fig Grape Rosehips
## NA 35.84000 33.63750 NA NA NA
## Strawberry Tomato
## 42.33333 NA## Calcul proportion change between G2 and G27
((tapply(data_sum$Nb_adults[data_sum$Generation=="27"],data_sum$Fruit_s[data_sum$Generation=="27"],mean) -
tapply(data_sum$Nb_adults[data_sum$Generation=="2"],data_sum$Fruit_s[data_sum$Generation=="2"],mean)) /
tapply(data_sum$Nb_adults[data_sum$Generation=="2"],data_sum$Fruit_s[data_sum$Generation=="2"],mean)) * 100## Blackcurrant Cherry Cranberry Fig Grape Rosehips
## NA 307.27273 73.50843 NA NA NA
## Strawberry Tomato
## 247.27919 NA#
# (abs(tapply(data_sum$fitness[data_sum$Generation=="27"],data_sum$Fruit_s[data_sum$Generation=="27"],mean) -
# tapply(data_sum$fitness[data_sum$Generation=="2"],data_sum$Fruit_s[data_sum$Generation=="2"],mean)) /
# abs(tapply(data_sum$fitness[data_sum$Generation=="2"],data_sum$Fruit_s[data_sum$Generation=="2"],mean))) * 100
## Calcul proportion change between G2 and phase2 (because G8 is not representative)
((tapply(data_sum$Nb_adults[data_sum$Generation=="8"],data_sum$Fruit_s[data_sum$Generation=="8"],mean) -
tapply(data_sum$Nb_adults[data_sum$Generation=="2"],data_sum$Fruit_s[data_sum$Generation=="2"],mean)) /
tapply(data_sum$Nb_adults[data_sum$Generation=="2"],data_sum$Fruit_s[data_sum$Generation=="2"],mean)) * 100## Blackcurrant Cherry Cranberry Fig Grape Rosehips
## NA 289.61039 85.02183 NA NA NA
## Strawberry Tomato
## 89.19658 NA## Compute proportion change between G2 and phase 2 (because G8 is not representative)
((tapply(data_sum$Nb_adults[data_sum$Phase=="pool"],data_sum$Fruit_s[data_sum$Phase=="pool"],mean) -
tapply(data_sum$Nb_adults[data_sum$Phase=="first_prepool"],data_sum$Fruit_s[data_sum$Phase=="first_prepool"],mean)) /
tapply(data_sum$Nb_adults[data_sum$Phase=="first_prepool"],data_sum$Fruit_s[data_sum$Phase=="first_prepool"],mean)) * 100## Blackcurrant Cherry Cranberry Fig Grape Rosehips
## NA 123.10237 93.62985 NA NA NA
## Strawberry Tomato
## 45.43089 NA((tapply(data_sum$Nb_adults[data_sum$Phase=="second_postpool"],data_sum$Fruit_s[data_sum$Phase=="second_postpool"],mean) -
tapply(data_sum$Nb_adults[data_sum$Phase=="first_prepool"],data_sum$Fruit_s[data_sum$Phase=="first_prepool"],mean)) /
tapply(data_sum$Nb_adults[data_sum$Phase=="first_prepool"],data_sum$Fruit_s[data_sum$Phase=="first_prepool"],mean)) * 100## Blackcurrant Cherry Cranberry Fig Grape Rosehips
## NA 230.1300 176.0760 NA NA NA
## Strawberry Tomato
## 145.1609 NA2.2 Analysis
######## Models
mod1 <- lme4::lmer(fitness ~ 1 + (1|Generation) + (1|Generation:Fruit_s),
weights = N, data = data_sum, REML = FALSE)
mod2 <- lme4::lmer(fitness~ Phase + (1|Generation) + (1|Generation:Fruit_s),
weights = N, data = data_sum, REML = FALSE)
mod3 <- lme4::lmer(fitness ~ Generation + (1|Generation) + (1|Generation:Fruit_s),
weights = N, data = data_sum, REML = FALSE)
mod4 <- lme4::lmer(fitness ~ Fruit_s + (1|Generation) + (1|Generation:Fruit_s),
weights = N, data = data_sum, REML = FALSE)
mod5 <- lme4::lmer(fitness ~ Fruit_s*Phase + (1|Generation) + (1|Generation:Fruit_s),
weights = N, data = data_sum, REML = FALSE)
mod6 <- lme4::lmer(fitness ~ Fruit_s*Generation + (1|Generation) + (1|Generation:Fruit_s),
weights = N, data = data_sum, REML = FALSE)
mod7 <- lme4::lmer(fitness ~ Fruit_s + Phase + (1|Generation) + (1|Generation:Fruit_s),
weights = N, data = data_sum, REML = FALSE)
mod8 <- lme4::lmer(fitness ~ Fruit_s + Generation + (1|Generation) + (1|Generation:Fruit_s),
weights = N, data = data_sum, REML = FALSE)
MuMIn::model.sel(mod1, mod2, mod3, mod4, mod5, mod6, mod7, mod8)## Model selection table
## (Int) Phs Gnr Frt_s Frt_s:Phs Frt_s:Gnr df logLik AICc delta
## mod7 -0.3309 + + 8 -122.149 260.9 0.00
## mod2 -0.4770 + 6 -126.389 265.1 4.23
## mod5 -0.5293 + + + 12 -120.494 266.3 5.42
## mod8 -0.1564 0.03884 + 7 -131.839 278.1 17.24
## mod6 -0.3216 0.04920 + + 9 -130.751 280.3 19.36
## mod3 -0.3093 0.03926 5 -136.100 282.4 21.55
## mod4 0.4371 + 6 -139.518 291.4 30.48
## mod1 0.2875 4 -143.957 296.1 35.18
## weight
## mod7 0.842
## mod2 0.102
## mod5 0.056
## mod8 0.000
## mod6 0.000
## mod3 0.000
## mod4 0.000
## mod1 0.000
## Models ranked by AICc(x)
## Random terms (all models):
## '1 | Generation', '1 | Generation:Fruit_s'summary(mod7)## Linear mixed model fit by maximum likelihood ['lmerMod']
## Formula:
## fitness ~ Fruit_s + Phase + (1 | Generation) + (1 | Generation:Fruit_s)
## Data: data_sum
## Weights: N
##
## AIC BIC logLik deviance df.resid
## 260.3 288.4 -122.1 244.3 239
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -5.2877 -0.3571 0.0899 0.4745 2.2266
##
## Random effects:
## Groups Name Variance Std.Dev.
## Generation:Fruit_s (Intercept) 0.05857 0.2420
## Generation (Intercept) 0.01257 0.1121
## Residual 2.11342 1.4538
## Number of obs: 247, groups: Generation:Fruit_s, 72; Generation, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) -0.33091 0.11980 -2.762
## Fruit_sCranberry -0.16282 0.08537 -1.907
## Fruit_sStrawberry -0.26491 0.08767 -3.022
## Phasepool 0.57826 0.14805 3.906
## Phasesecond_postpool 0.99779 0.11899 8.386
##
## Correlation of Fixed Effects:
## (Intr) Frt_sC Frt_sS Phaspl
## Frt_sCrnbrr -0.385
## Frt_sStrwbr -0.381 0.532
## Phasepool -0.641 -0.025 -0.025
## Phsscnd_pst -0.817 0.003 0.011 0.658mod_Phase <- lme4::lmer(fitness ~ Fruit_s + Phase + (1|Generation) + (1|Generation:Fruit_s),
weights = N, data = data_sum)
#Posthoc
emmeans::emmeans(mod_Phase, list(pairwise ~ Phase), adjust = "tukey") ### $`emmeans of Phase`
## Phase emmean SE df lower.CL upper.CL
## first_prepool -0.474 0.1179 159 -0.758 -0.189
## pool 0.106 0.1088 34492 -0.154 0.366
## second_postpool 0.524 0.0549 543 0.392 0.655
##
## Results are averaged over the levels of: Fruit_s
## Degrees-of-freedom method: kenward-roger
## Confidence level used: 0.95
## Conf-level adjustment: sidak method for 3 estimates
##
## $`pairwise differences of Phase`
## contrast estimate SE df t.ratio p.value
## first_prepool - pool -0.580 0.160 527 -3.617 0.0010
## first_prepool - second_postpool -0.997 0.130 190 -7.671 <.0001
## pool - second_postpool -0.418 0.122 6553 -3.429 0.0018
##
## Results are averaged over the levels of: Fruit_s
## Degrees-of-freedom method: kenward-roger
## P value adjustment: tukey method for comparing a family of 3 estimatesemmeans::emmeans(mod_Phase, list(pairwise ~ Phase+Fruit_s), adjust = "tukey") ### $`emmeans of Phase, Fruit_s`
## Phase Fruit_s emmean SE df lower.CL upper.CL
## first_prepool Cherry -0.3309 0.1296 232 -0.6926 0.0308
## pool Cherry 0.2490 0.1231 27016 -0.0915 0.5894
## second_postpool Cherry 0.6665 0.0760 2263 0.4560 0.8770
## first_prepool Cranberry -0.4941 0.1272 223 -0.8493 -0.1389
## pool Cranberry 0.0858 0.1179 32019 -0.2403 0.4118
## second_postpool Cranberry 0.5033 0.0728 901 0.3015 0.7051
## first_prepool Strawberry -0.5961 0.1283 226 -0.9543 -0.2378
## pool Strawberry -0.0162 0.1190 34230 -0.3454 0.3130
## second_postpool Strawberry 0.4014 0.0754 2177 0.1926 0.6101
##
## Degrees-of-freedom method: kenward-roger
## Confidence level used: 0.95
## Conf-level adjustment: sidak method for 9 estimates
##
## $`pairwise differences of Phase, Fruit_s`
## contrast estimate SE df
## first_prepool,Cherry - pool,Cherry -0.580 0.1603 527
## first_prepool,Cherry - second_postpool,Cherry -0.997 0.1300 190
## first_prepool,Cherry - first_prepool,Cranberry 0.163 0.0877 2089
## first_prepool,Cherry - pool,Cranberry -0.417 0.1810 822
## first_prepool,Cherry - second_postpool,Cranberry -0.834 0.1573 365
## first_prepool,Cherry - first_prepool,Strawberry 0.265 0.0898 3565
## first_prepool,Cherry - pool,Strawberry -0.315 0.1820 869
## first_prepool,Cherry - second_postpool,Strawberry -0.732 0.1588 424
## pool,Cherry - second_postpool,Cherry -0.418 0.1218 6553
## pool,Cherry - first_prepool,Cranberry 0.743 0.1845 936
## pool,Cherry - pool,Cranberry 0.163 0.0877 2089
## pool,Cherry - second_postpool,Cranberry -0.254 0.1526 5336
## pool,Cherry - first_prepool,Strawberry 0.845 0.1855 953
## pool,Cherry - pool,Strawberry 0.265 0.0898 3565
## pool,Cherry - second_postpool,Strawberry -0.152 0.1542 9464
## second_postpool,Cherry - first_prepool,Cranberry 1.161 0.1565 449
## second_postpool,Cherry - pool,Cranberry 0.581 0.1475 13466
## second_postpool,Cherry - second_postpool,Cranberry 0.163 0.0877 2089
## second_postpool,Cherry - first_prepool,Strawberry 1.263 0.1573 422
## second_postpool,Cherry - pool,Strawberry 0.683 0.1484 10341
## second_postpool,Cherry - second_postpool,Strawberry 0.265 0.0898 3565
## first_prepool,Cranberry - pool,Cranberry -0.580 0.1603 527
## first_prepool,Cranberry - second_postpool,Cranberry -0.997 0.1300 190
## first_prepool,Cranberry - first_prepool,Strawberry 0.102 0.0861 2174
## first_prepool,Cranberry - pool,Strawberry -0.478 0.1819 882
## first_prepool,Cranberry - second_postpool,Strawberry -0.895 0.1563 440
## pool,Cranberry - second_postpool,Cranberry -0.418 0.1218 6553
## pool,Cranberry - first_prepool,Strawberry 0.682 0.1820 848
## pool,Cranberry - pool,Strawberry 0.102 0.0861 2174
## pool,Cranberry - second_postpool,Strawberry -0.316 0.1495 13058
## second_postpool,Cranberry - first_prepool,Strawberry 1.099 0.1556 356
## second_postpool,Cranberry - pool,Strawberry 0.520 0.1487 5535
## second_postpool,Cranberry - second_postpool,Strawberry 0.102 0.0861 2174
## first_prepool,Strawberry - pool,Strawberry -0.580 0.1603 527
## first_prepool,Strawberry - second_postpool,Strawberry -0.997 0.1300 190
## pool,Strawberry - second_postpool,Strawberry -0.418 0.1218 6553
## t.ratio p.value
## -3.617 0.0098
## -7.671 <.0001
## 1.860 0.6416
## -2.302 0.3420
## -5.304 <.0001
## 2.952 0.0772
## -1.729 0.7284
## -4.612 0.0002
## -3.429 0.0177
## 4.027 0.0020
## 1.860 0.6416
## -1.666 0.7672
## 4.554 0.0002
## 2.952 0.0772
## -0.988 0.9870
## 7.418 <.0001
## 3.937 0.0027
## 1.860 0.6416
## 8.027 <.0001
## 4.602 0.0001
## 2.952 0.0772
## -3.617 0.0098
## -7.671 <.0001
## 1.184 0.9599
## -2.627 0.1770
## -5.729 <.0001
## -3.429 0.0177
## 3.746 0.0060
## 1.184 0.9599
## -2.110 0.4661
## 7.065 <.0001
## 3.493 0.0142
## 1.184 0.9599
## -3.617 0.0098
## -7.671 <.0001
## -3.429 0.0177
##
## Degrees-of-freedom method: kenward-roger
## P value adjustment: tukey method for comparing a family of 9 estimates2.3 Plot
#Position dodge
pd <- ggplot2::position_dodge(0.3) # move them .05 to the left and right
#Extract slope and intercept
dat_predict_allfruits <- expand.grid(Generation=as.numeric(levels(as.factor(data_sum$Generation))),
Fruit_s=unique(data_sum$Fruit_s))
dat_predict_allfruits$Phase <- ifelse(dat_predict_allfruits$Generation<= 7, "first_prepool",
ifelse(dat_predict_allfruits$Generation>=12, "second_postpool", "pool"))
dat_predict_allfruits$fitness_predicted <- predict(mod_Phase, newdata = dat_predict_allfruits,
re.form= NA, type = "response")
#REAL DATA
#Add G1 G6 and G7
TEMP_lineG1 <- c(rep(NA, 2), 1,rep(NA, 5))
TEMP_lineG6 <- c(rep(NA, 2), 6,rep(NA, 5))
TEMP_lineG7 <- c(rep(NA, 2), 7,rep(NA, 5))
TEMP_total <- rbind(data_sum,TEMP_lineG1,TEMP_lineG1,TEMP_lineG1,TEMP_lineG6,TEMP_lineG6,TEMP_lineG6,TEMP_lineG7,TEMP_lineG7,TEMP_lineG7)
TEMP_total$Fruit_s[TEMP_total$Generation == "6"] <- c("Strawberry", "Cranberry", "Cherry")
TEMP_total$Fruit_s[TEMP_total$Generation == "1"] <- c("Strawberry", "Cranberry", "Cherry")
TEMP_total$Fruit_s[TEMP_total$Generation == "7"] <- c("Strawberry", "Cranberry", "Cherry")
tail(TEMP_total)## Line Fruit_s Generation Phase N Nb_adults sd fitness se_fitness
## 251 <NA> Strawberry 6 <NA> NA NA NA NA NA
## 252 <NA> Cranberry 6 <NA> NA NA NA NA NA
## 253 <NA> Cherry 6 <NA> NA NA NA NA NA
## 254 <NA> Strawberry 7 <NA> NA NA NA NA NA
## 255 <NA> Cranberry 7 <NA> NA NA NA NA NA
## 256 <NA> Cherry 7 <NA> NA NA NA NA NA## Add label
TEMP_anno <- data.frame(x1 = c(3.5, 3.5, 10, 3.5, 3.5, 10, 3.5, 3.5, 10),
x2 = c(9, 17.5, 17.5,9, 17.5, 17.5,9,17.5, 17.5),
y1 = c(1,2,1.2,1,2,1.2,1,2,1.2),
y2 = c(1.25, 2.25, 1.45, 1.25, 2.25, 1.45,1.25, 2.25, 1.45),
xstar = c(6.5,10,14,6.5,10,14,6.5,10,14),
ystar = c(1.5,2.5,1.7,1.5,2.5,1.7,1.5,2.5,1.7),
lab = c("**", "***", "**", "**", "***", "**", "**" ,"***", "**"),
Fruit_s = c("Cherry", "Cherry", "Cherry",
"Cranberry", "Cranberry", "Cranberry",
"Strawberry","Strawberry","Strawberry"),
Line = NA)
TEMP_anno## x1 x2 y1 y2 xstar ystar lab Fruit_s Line
## 1 3.5 9.0 1.0 1.25 6.5 1.5 ** Cherry NA
## 2 3.5 17.5 2.0 2.25 10.0 2.5 *** Cherry NA
## 3 10.0 17.5 1.2 1.45 14.0 1.7 ** Cherry NA
## 4 3.5 9.0 1.0 1.25 6.5 1.5 ** Cranberry NA
## 5 3.5 17.5 2.0 2.25 10.0 2.5 *** Cranberry NA
## 6 10.0 17.5 1.2 1.45 14.0 1.7 ** Cranberry NA
## 7 3.5 9.0 1.0 1.25 6.5 1.5 ** Strawberry NA
## 8 3.5 17.5 2.0 2.25 10.0 2.5 *** Strawberry NA
## 9 10.0 17.5 1.2 1.45 14.0 1.7 ** Strawberry NATEMP_title <- data.frame(xtitle = c(14, 14, 14),
ytitle = c(3, 3, 3),
title = c("Cherry", "Cranberry", "Strawberry"),
Fruit_s = c("Cherry", "Cranberry", "Strawberry"),
Line = NA)
TEMP_title## xtitle ytitle title Fruit_s Line
## 1 14 3 Cherry Cherry NA
## 2 14 3 Cranberry Cranberry NA
## 3 14 3 Strawberry Strawberry NAPLOT_FITNESS_CHERRY <- ggplot(data = TEMP_total[TEMP_total$Fruit_s == "Cherry",],
aes(x = factor(Generation),group = Line, y = fitness, colour =Fruit_s)) +
geom_errorbar(aes(ymin =fitness-1.96*se_fitness, ymax = fitness + 1.96*se_fitness),
width=.1,position = pd, size = 0.2,color = "black") +
geom_line(size = 0.3,position = pd) +
geom_line(data = dat_predict_allfruits[dat_predict_allfruits$Fruit_s == "Cherry",],
aes(x = factor(Generation), y = fitness_predicted,
colour = "black", group = Phase), size = 0.5) +
geom_point(size =1, position = pd, shape =21, fill = "white") +
ylim(-3, 3.05) +
ylab("Fitness") +
xlab("Generation") +
geom_text(data = TEMP_anno[TEMP_anno$Fruit_s == "Cherry",], aes(x = xstar, y = ystar, label = lab), size =3.3) +
geom_segment(data = TEMP_anno[TEMP_anno$Fruit_s == "Cherry",], aes(x = x1, xend = x1,
y = y1, yend = y2), size = 0.4) +
geom_segment(data = TEMP_anno[TEMP_anno$Fruit_s == "Cherry",], aes(x = x2, xend = x2,
y = y1, yend = y2), size = 0.4) +
geom_segment(data = TEMP_anno[TEMP_anno$Fruit_s == "Cherry",], aes(x = x1, xend = x2,
y = y2, yend = y2), size = 0.4) +
scale_color_manual(values = c("black", "#BC3C6D", "#FDB424", "#3FAA96")) +
ggtitle("Cherry") +
theme_LO_adaptation + theme(plot.title = element_text(color = "#BC3C6D"))
PLOT_FITNESS_CHERRY
PLOT_FITNESS_CRANB <- ggplot2::ggplot(data = TEMP_total[TEMP_total$Fruit_s == "Cranberry",],
aes(x = factor(Generation),group = Line, y = fitness, colour =Fruit_s)) +
geom_errorbar(aes(ymin =fitness-1.96*se_fitness, ymax = fitness + 1.96*se_fitness),
width=.1,position = pd, size = 0.2,color = "black") +
geom_line(size = 0.3,position = pd) +
geom_line(data = dat_predict_allfruits[dat_predict_allfruits$Fruit_s == "Cranberry",], aes(x = factor(Generation), y = fitness_predicted, colour = "black", group = Phase),
size = 0.5) +
geom_point(size =1, position = pd, shape =21, fill = "white") +
ylim(-3, 3.05) +
ylab("Fitness") +
xlab("Generation") +
geom_text(data = TEMP_anno[TEMP_anno$Fruit_s == "Cranberry",], aes(x = xstar, y = ystar, label = lab), size =3.3) +
geom_segment(data = TEMP_anno[TEMP_anno$Fruit_s == "Cranberry",], aes(x = x1, xend = x1,
y = y1, yend = y2), size = 0.4) +
geom_segment(data = TEMP_anno[TEMP_anno$Fruit_s == "Cranberry",], aes(x = x2, xend = x2,
y = y1, yend = y2), size = 0.4) +
geom_segment(data = TEMP_anno[TEMP_anno$Fruit_s == "Cranberry",], aes(x = x1, xend = x2,
y = y2, yend = y2), size = 0.4) +
scale_color_manual(values = c("black", "#FDB424")) +
ggtitle("Cranberry") +
theme_LO_adaptation + theme(plot.title = element_text(color = "#FDB424"))
PLOT_FITNESS_CRANB
PLOT_FITNESS_STRAW <- ggplot2::ggplot(data = TEMP_total[TEMP_total$Fruit_s == "Strawberry",],
aes(x = factor(Generation),group = Line, y = fitness, colour =Fruit_s)) +
geom_errorbar(aes(ymin =fitness-1.96*se_fitness, ymax = fitness + 1.96*se_fitness),
width=.1,position = pd, size = 0.2,color = "black") +
geom_line(size = 0.3,position = pd) +
geom_line(data = dat_predict_allfruits[dat_predict_allfruits$Fruit_s == "Strawberry",],
aes(x = factor(Generation), y = fitness_predicted, colour = "black", group = Phase),
size = 0.5) +
geom_point(size =1, position = pd, shape =21, fill = "white") +
ylim(-3, 3.05) +
ylab("Fitness") +
xlab("Generation") +
geom_text(data = TEMP_anno[TEMP_anno$Fruit_s == "Strawberry",],
aes(x = xstar, y = ystar, label = lab), size =3.3) +
geom_segment(data = TEMP_anno[TEMP_anno$Fruit_s == "Strawberry",],
aes(x = x1, xend = x1, y = y1, yend = y2), size = 0.4) +
geom_segment(data = TEMP_anno[TEMP_anno$Fruit_s == "Strawberry",],
aes(x = x2, xend = x2, y = y1, yend = y2), size = 0.4) +
geom_segment(data = TEMP_anno[TEMP_anno$Fruit_s == "Strawberry",],
aes(x = x1, xend = x2, y = y2, yend = y2), size = 0.4) +
scale_color_manual(values = c("black", "#3FAA96")) +
ggtitle("Strawberry") +
theme_LO_adaptation + theme(plot.title = element_text(color = "#3FAA96"))
PLOT_FITNESS_STRAW
DYNAMIQUE_THREE_JOIN <- cowplot::ggdraw() +
cowplot::draw_plot(PLOT_FITNESS_CHERRY + theme(axis.title.x = element_blank(),
axis.text.x = element_blank(),
axis.ticks.x = element_blank(),
axis.line.x = element_blank()),
x = 0, y = 0.66, width = 1, height = 0.33) +
cowplot::draw_plot(PLOT_FITNESS_CRANB + theme(axis.title.x = element_blank(),
axis.text.x = element_blank(),
axis.ticks.x = element_blank(),
axis.line.x = element_blank()),
x = 0, y = 0.33, width = 1, height = 0.33) +
cowplot::draw_plot(PLOT_FITNESS_STRAW,
x = 0, y = 00, width = 1, height = 0.33) +
cowplot::draw_plot_label(c("A", "B", "C"),
x = c(0, 0, 0),
y = c(1, 0.66, 0.33),
hjust = c(-0.5, -0.5, -0.5),
vjust = c(1.5, 1.5, 1.5),
size = 12)
DYNAMIQUE_THREE_JOIN
cowplot::save_plot(file =here::here("figures", "FIG_Adaptation.pdf"), DYNAMIQUE_THREE_JOIN, base_height = 17/cm(1), base_width = 11/cm(1), dpi = 1200)3 HETEROGENEITY
3.1 Plot fitness
pd <- position_dodge(width = 0.5)
################### INTERMEDIATE PHENOTYPING
# Re-order levels of Line
data_sum_G7 <- data_logchange[data_logchange$Generation == "7",]
data_sum_G7$Line <- factor(data_sum_G7$Line, levels= c("CE2", "CE1", "CE4", "CE3",
"CR2", "CR3", "CR5", "CR1", "CR4",
"FR2", "FR3", "FR5", "FR1", "FR4"))
#Plot:
symp_allop_g7 <- ggplot(data = data_sum_G7,
aes(x = Line, y = logchange, group = Treatment,
color = Fruit_s, shape = Treatment, fill = Fruit_s)) +
geom_hline(yintercept = 0, linetype = "dashed", color = "grey", size = 0.5) +
geom_errorbar(aes(ymin = logchange-1.96*sd_logchange,
ymax =logchange + 1.96*sd_logchange, linetype = Line_type),
width = 0.2, size = 0.5, alpha = 0.6,position = pd) +
geom_point(position = pd, fill = "white", size =3) +
ylab("Fitness change between intermediate\nand initial phenotyping steps") +
xlab("Populations") +
labs(shape = "Test fruit", color = "Selection fruit") +
guides(color = FALSE,
shape = guide_legend(override.aes = list(fill = c("black")))) +
scale_shape_manual(values = c(21, 22, 24)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
scale_fill_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
theme_LO_sober + theme(axis.text.x = element_blank()) +
coord_cartesian(expand = FALSE, ylim = c(-2, 2), xlim=c(0, 15),clip = "off") +
ggtitle("Intermediate phenotyping step") +
annotate("text", x = 2.5, y =1.5, label = 'bold(" Evolved\non cherry")',
size =3,colour = "#BC3C6D",parse = TRUE) +
annotate("text", x = 7.5, y =1.5, label = 'bold(" Evolved\non cranberry")',
size =3,colour = "#FDB424",parse = TRUE) +
annotate("text", x = 12.5, y =1.5, label = 'bold(" Evolved\non strawberry")',
size =3,colour = "#3FAA96",parse = TRUE) +
geom_segment(x = 15.5, y = -0.1, xend= 15.5, yend = -1.2, size = 0.15,
arrow = arrow(length = unit(0.05, "npc")),colour = "black") +
geom_segment(x = 15.5, y = 0.1, xend= 15.5, yend = 1.2, size = 0.15,
arrow = arrow(length = unit(0.05, "npc")),colour = "black") +
annotate("text", x = 16.2, y = 0.5, label = 'bold(" Fitness\nincrease")',
size =3,colour = "black",parse = TRUE, angle =90) +
annotate("text", x = 16.2, y =-0.5, label = 'bold(" Fitness\ndecrease")',
size =3,colour = "black",parse = TRUE, angle =90)
symp_allop_g7
################### FINAL PHENOTYPING
# Re-order levels of Line
data_sum_G29 <- data_logchange[data_logchange$Generation == "29",]
data_sum_G29$Line <- factor(data_sum_G29$Line, levels = c("CEA", "CEB", "CEC",
"CRD", "CRA", "CRC", "CRB", "CRE",
"FRA", "FRC", "FRB"))
symp_allop_G29 <- ggplot(data = data_sum_G29,
aes(x = Line, y=logchange, group = Treatment,
color =Fruit_s, shape = Treatment,fill =Fruit_s)) +
geom_hline(yintercept = 0, linetype = "dashed", color = "grey", size = 0.5) +
geom_point(size =3, position = pd) +
geom_errorbar(aes(ymin =logchange-1.96*sd_logchange, ymax =logchange + 1.96*sd_logchange),
width= 0.2, size = 0.5, alpha = 0.6,position = pd) +
ylab("Fitness change between final\nand initial phenotyping steps") +
xlab("Populations") +
labs(shape = "Test fruit", color = "Selection fruit") +
guides(fill = FALSE, alpha = FALSE) +
scale_shape_manual(values = c(21, 22, 24)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
scale_fill_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
theme_LO_sober + theme(axis.text.x = element_blank()) +
coord_cartesian(expand = FALSE, ylim = c(-2, 2), xlim=c(0, 11),clip = "off") +
annotate("text", x = 2, y =1.5, label = 'bold(" Evolved\non cherry")',
size =3,colour = "#BC3C6D",parse = TRUE) +
annotate("text", x = 6, y =1.5, label = 'bold(" Evolved\non cranberry")',
size =3,colour = "#FDB424",parse = TRUE) +
annotate("text", x = 10, y =1.5, label = 'bold(" Evolved\non strawberry")',
size =3,colour = "#3FAA96",parse = TRUE) +
geom_segment(x = 11.5, y = -0.1, xend= 11.5, yend = -1.2, size = 0.15,
arrow = arrow(length = unit(0.05, "npc")),colour = "black") +
geom_segment(x = 11.5, y = 0.1, xend= 11.5, yend = 1.2, size = 0.15,
arrow = arrow(length = unit(0.05, "npc")),colour = "black") +
annotate("text", x = 12, y = 0.5, label = 'bold(" Fitness\nincrease")',
size =3,colour = "black",parse = TRUE, angle =90) +
annotate("text", x = 12, y = -0.5, label = 'bold(" Fitness\ndecrease")',
size =3,colour = "black", parse = TRUE, angle =90) +
ggtitle("Final phenotyping step")
symp_allop_G29
### Add stroke
DIF_FITNESS_G7 <- symp_allop_g7 +
geom_point(aes(alpha = SA, color = interaction(SA,Fruit_s)), position = pd, size =3, stroke =1.5, fill = "white") +
scale_alpha_manual(values = c(0, 1)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96", "#7C2748", "#CA8702", "#328677", "#BC3C6D", "#FDB424", "#3FAA96"))
DIF_FITNESS_G7
DIF_FITNESS_G29 <- symp_allop_G29 + geom_point(aes(alpha = SA, color = interaction(SA,Fruit_s)), position = pd, size =3, stroke =1.5) +
scale_alpha_manual(values = c(0, 1)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96", "#7C2748", "#CA8702", "#328677", "#BC3C6D", "#FDB424", "#3FAA96"))
DIF_FITNESS_G29
plot_legend <- ggplot(data = data_logchange, aes(x = Line, y=logchange, group = Treatment,
color =Fruit_s, shape = Treatment,fill =Fruit_s)) +
geom_point(size =3, position = pd) +
labs(shape = "Test fruit", color = "Selection fruit") +
scale_shape_manual(values = c(21, 22, 24)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
theme_LO_sober +
guides(fill = FALSE) +
theme(legend.key.size = unit(0.5, "cm"),
axis.text.x = element_blank()) +
coord_cartesian(expand = FALSE, ylim = c(-2, 2), xlim=c(0, 11),clip = "off")
###########################################################
#############################################ALL PLOT
legend_trade <- lemon::g_legend(plot_legend)
SYMP_ALLOP_TOTAL <- cowplot::ggdraw() +
cowplot::draw_plot(DIF_FITNESS_G7 + theme(legend.position = 'none',
plot.margin =unit(c(0.5, 3.5, 0.5, 0.5), "cm")),
x = 0, y = 0.5, width = 1, height = 0.5) +
cowplot::draw_plot(DIF_FITNESS_G29 + theme(legend.position = 'none',
plot.margin =unit(c(0.5, 3.5, 0.5, 0.5), "cm")),
x = 0, y = 0, width = 1, height = 0.5) +
cowplot::draw_plot_label(c("A", "B"),
x = c(0, 0), y = c(1, 0.5),
size = 14) +
cowplot::draw_plot(legend_trade, x = 0.93, y = 0.5, width = 0.001, height = 0.001)
SYMP_ALLOP_TOTAL
cowplot::save_plot(file =here::here("figures", "FIG_Heterogeneity.pdf"),
SYMP_ALLOP_TOTAL, base_height = 20/cm(1), base_width = 20/cm(1), dpi = 1200)4 Traits
4.1 Correlation between Nb_eggs and emergence_rate
tapply(data_G0$Nb_eggs, data_G0$Treatment, mean)## Cherry Cranberry Strawberry
## 149.49 122.85 107.84diag(tapply(data_G7$Nb_eggs, list(data_G7$Treatment,data_G7$Fruit_s), mean))## Cherry Cranberry Strawberry
## 195.4375 181.1591 174.2923diag(tapply(data_G29$Nb_eggs, list(data_G29$Treatment,data_G29$Fruit_s), mean))## Cherry Cranberry Strawberry
## 103.7667 123.7000 126.7333tapply(data_G0$Emergence_rate, data_G0$Treatment, mean)## Cherry Cranberry Strawberry
## 0.1837216 0.2660867 0.2777895diag(tapply(data_G7$Emergence_rate, list(data_G7$Treatment,data_G7$Fruit_s), mean))## Cherry Cranberry Strawberry
## 0.1816565 0.2070999 0.2728989diag(tapply(data_G29$Emergence_rate, list(data_G29$Treatment,data_G29$Fruit_s), mean))## Cherry Cranberry Strawberry
## 0.3421827 0.2794123 0.3543303#Plot
Nb_eggsEmergence_rateG7 <- ggplot(data = data_logchangeNb_eggsEmergence_rate[data_logchangeNb_eggsEmergence_rate$Generation==7,],
aes(x = Nb_eggslogchange,y = Emergence_ratelogchange, group = Treatment, shape = Treatment)) +
geom_errorbarh(aes(xmin = Nb_eggslogchange-1.96*sdNb_eggslogchange,
xmax = Nb_eggslogchange+1.96*sdNb_eggslogchange),
height=0.02,size=0.2,alpha=1) +
geom_errorbar(aes(ymin = Emergence_ratelogchange-1.96*sdEmergence_ratelogchange,
ymax = Emergence_ratelogchange+1.96*sdEmergence_ratelogchange),
width=0.02,size=0.2,alpha=1) +
geom_point(aes(shape=Treatment, color=Treatment)) +
xlab("Change in number of eggs between\nintermediate and initial phenotyping steps") +
ylab("Change in egg-to-adult between\nintermediate and initial phenotyping steps") +
xlim(-0.75, 1) +
ylim(-1.75, 1.5) +
labs(shape = "Test fruit") +
scale_shape_manual(values = c(21, 22, 24)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
scale_fill_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
theme_LO_sober
Nb_eggsEmergence_rateG7 <- Nb_eggsEmergence_rateG7 + facet_wrap(~ Fruit_s)
Nb_eggsEmergence_rateG29 <- ggplot(data = data_logchangeNb_eggsEmergence_rate[data_logchangeNb_eggsEmergence_rate$Generation==29,],
aes(x = Nb_eggslogchange,y = Emergence_ratelogchange, group = Treatment, shape = Treatment)) +
geom_errorbarh(aes(xmin = Nb_eggslogchange-1.96*sdNb_eggslogchange,
xmax = Nb_eggslogchange+1.96*sdNb_eggslogchange),
height=0.02,size=0.2,alpha=1) +
geom_errorbar(aes(ymin = Emergence_ratelogchange-1.96*sdEmergence_ratelogchange,
ymax = Emergence_ratelogchange+1.96*sdEmergence_ratelogchange),
width=0.02,size=0.2,alpha=1) +
geom_point(aes(shape=Treatment, color=Treatment)) +
xlab("Change in number of eggs between\nfinal and initial phenotyping steps") +
ylab("Change in egg-to-adult viability between\nfinal and initial phenotyping steps") +
xlim(-0.75, 1) +
ylim(-1.75, 1.5) +
labs(shape = "Test fruit", color = "Test fruit") +
scale_shape_manual(values = c(21, 22, 24)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
scale_fill_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
theme_LO_sober
Nb_eggsEmergence_rateG29 <- Nb_eggsEmergence_rateG29 + facet_wrap(~ Fruit_s)
legend_trade <- lemon::g_legend(Nb_eggsEmergence_rateG29)
Nb_eggsEmergence_rate <- cowplot::ggdraw() +
cowplot::draw_plot(Nb_eggsEmergence_rateG7 + theme(legend.position = 'none',
plot.margin =unit(c(0.5, 3.5, 0.5, 0.5), "cm")),
x = 0, y = 0.5, width = 1, height = 0.5) +
cowplot::draw_plot(Nb_eggsEmergence_rateG29 + theme(legend.position = 'none',
plot.margin =unit(c(0.5, 3.5, 0.5, 0.5), "cm")),
x = 0, y = 0, width = 1, height = 0.5) +
cowplot::draw_plot_label(c("A", "B"),
x = c(0, 0), y = c(1, 0.5),
size = 14) +
cowplot::draw_plot(legend_trade, x = 0.93, y = 0.5, width = 0.001, height = 0.001)
cowplot::save_plot(file =here::here("figures", "FIG_SX_Nb_eggsEmergence_rate.pdf"),
Nb_eggsEmergence_rate, base_height = 20/cm(1), base_width = 20/cm(1), dpi = 1200)4.2 Plot Nb_eggs
pd <- position_dodge(width = 0.5)
################### INTERMEDIATE PHENOTYPING
# Re-order levels of Line
data_sum_G7 <- data_logchangeNb_eggs[data_logchangeNb_eggs$Generation == "7",]
data_sum_G7$Line <- factor(data_sum_G7$Line, levels= c("CE2", "CE1", "CE4", "CE3",
"CR2", "CR3", "CR5", "CR1", "CR4",
"FR2", "FR3", "FR5", "FR1", "FR4"))
#Plot:
symp_allop_g7 <- ggplot(data = data_sum_G7,
aes(x = Line, y = logchange, group = Treatment,
color = Fruit_s, shape = Treatment, fill = Fruit_s)) +
geom_hline(yintercept = 0, linetype = "dashed", color = "grey", size = 0.5) +
geom_errorbar(aes(ymin = logchange-1.96*sd_logchange,
ymax =logchange + 1.96*sd_logchange, linetype = Line_type),
width = 0.2, size = 0.5, alpha = 0.6,position = pd) +
geom_point(position = pd, fill = "white", size =3) +
ylab("Change in number of eggs between\nintermediate and initial phenotyping steps") +
xlab("Populations") +
labs(shape = "Test fruit", color = "Selection fruit") +
guides(color = FALSE,
shape = guide_legend(override.aes = list(fill = c("black")))) +
scale_shape_manual(values = c(21, 22, 24)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
scale_fill_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
theme_LO_sober + theme(axis.text.x = element_blank()) +
coord_cartesian(expand = FALSE, ylim = c(-2, 2), xlim=c(0, 15),clip = "off") +
ggtitle("Intermediate phenotyping step") +
annotate("text", x = 2.5, y =1.5, label = 'bold(" Evolved\non cherry")',
size =3,colour = "#BC3C6D",parse = TRUE) +
annotate("text", x = 7.5, y =1.5, label = 'bold(" Evolved\non cranberry")',
size =3,colour = "#FDB424",parse = TRUE) +
annotate("text", x = 12.5, y =1.5, label = 'bold(" Evolved\non strawberry")',
size =3,colour = "#3FAA96",parse = TRUE)
symp_allop_g7
################### FINAL PHENOTYPING
# Re-order levels of Line
data_sum_G29 <- data_logchangeNb_eggs[data_logchangeNb_eggs$Generation == "29",]
data_sum_G29$Line <- factor(data_sum_G29$Line, levels = c("CEA", "CEB", "CEC",
"CRD", "CRA", "CRC", "CRB", "CRE",
"FRA", "FRC", "FRB"))
symp_allop_G29 <- ggplot(data = data_sum_G29,
aes(x = Line, y=logchange, group = Treatment,
color =Fruit_s, shape = Treatment,fill =Fruit_s)) +
geom_hline(yintercept = 0, linetype = "dashed", color = "grey", size = 0.5) +
geom_point(size =3, position = pd) +
geom_errorbar(aes(ymin =logchange-1.96*sd_logchange, ymax =logchange + 1.96*sd_logchange),
width= 0.2, size = 0.5, alpha = 0.6,position = pd) +
ylab("Change in number of eggs between\nfinal and initial phenotyping steps") +
xlab("Populations") +
labs(shape = "Test fruit", color = "Selection fruit") +
guides(fill = FALSE, alpha = FALSE) +
scale_shape_manual(values = c(21, 22, 24)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
scale_fill_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
theme_LO_sober + theme(axis.text.x = element_blank()) +
coord_cartesian(expand = FALSE, ylim = c(-2, 2), xlim=c(0, 11),clip = "off") +
annotate("text", x = 2, y =1.5, label = 'bold(" Evolved\non cherry")',
size =3,colour = "#BC3C6D",parse = TRUE) +
annotate("text", x = 6, y =1.5, label = 'bold(" Evolved\non cranberry")',
size =3,colour = "#FDB424",parse = TRUE) +
annotate("text", x = 10, y =1.5, label = 'bold(" Evolved\non strawberry")',
size =3,colour = "#3FAA96",parse = TRUE) +
ggtitle("Final phenotyping step")
symp_allop_G29
### Add stroke
DIF_FITNESS_G7 <- symp_allop_g7 +
geom_point(aes(alpha = SA, color = interaction(SA,Fruit_s)), position = pd, size =3, stroke =1.5, fill = "white") +
scale_alpha_manual(values = c(0, 1)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96", "#7C2748", "#CA8702", "#328677", "#BC3C6D", "#FDB424", "#3FAA96"))
DIF_FITNESS_G7
DIF_FITNESS_G29 <- symp_allop_G29 + geom_point(aes(alpha = SA, color = interaction(SA,Fruit_s)), position = pd, size =3, stroke =1.5) +
scale_alpha_manual(values = c(0, 1)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96", "#7C2748", "#CA8702", "#328677", "#BC3C6D", "#FDB424", "#3FAA96"))
DIF_FITNESS_G29
plot_legend <- ggplot(data = data_logchange, aes(x = Line, y=logchange, group = Treatment,
color =Fruit_s, shape = Treatment,fill =Fruit_s)) +
geom_point(size =3, position = pd) +
labs(shape = "Test fruit", color = "Selection fruit") +
scale_shape_manual(values = c(21, 22, 24)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
theme_LO_sober +
guides(fill = FALSE) +
theme(legend.key.size = unit(0.5, "cm"),
axis.text.x = element_blank()) +
coord_cartesian(expand = FALSE, ylim = c(-2, 2), xlim=c(0, 11),clip = "off")
###########################################################
#############################################ALL PLOT
legend_trade <- lemon::g_legend(plot_legend)
SYMP_ALLOP_TOTAL <- cowplot::ggdraw() +
cowplot::draw_plot(DIF_FITNESS_G7 + theme(legend.position = 'none',
plot.margin =unit(c(0.5, 3.5, 0.5, 0.5), "cm")),
x = 0, y = 0.5, width = 1, height = 0.5) +
cowplot::draw_plot(DIF_FITNESS_G29 + theme(legend.position = 'none',
plot.margin =unit(c(0.5, 3.5, 0.5, 0.5), "cm")),
x = 0, y = 0, width = 1, height = 0.5) +
cowplot::draw_plot_label(c("A", "B"),
x = c(0, 0), y = c(1, 0.5),
size = 14) +
cowplot::draw_plot(legend_trade, x = 0.93, y = 0.5, width = 0.001, height = 0.001)
DIF_FITNESS_G7Nb_eggs <- DIF_FITNESS_G7
DIF_FITNESS_G29Nb_eggs <- DIF_FITNESS_G29
cowplot::save_plot(file =here::here("figures", "FIG_SX_HeterogeneityNb_eggs.pdf"),
SYMP_ALLOP_TOTAL, base_height = 20/cm(1), base_width = 20/cm(1), dpi = 1200)4.3 Plot egg-to-adult viability
pd <- position_dodge(width = 0.5)
################### INTERMEDIATE PHENOTYPING
# Re-order levels of Line
data_sum_G7 <- data_logchangeEmergence_rate[data_logchangeEmergence_rate$Generation == "7",]
data_sum_G7$Line <- factor(data_sum_G7$Line, levels= c("CE2", "CE1", "CE4", "CE3",
"CR2", "CR3", "CR5", "CR1", "CR4",
"FR2", "FR3", "FR5", "FR1", "FR4"))
#Plot:
symp_allop_g7 <- ggplot(data = data_sum_G7,
aes(x = Line, y = logchange, group = Treatment,
color = Fruit_s, shape = Treatment, fill = Fruit_s)) +
geom_hline(yintercept = 0, linetype = "dashed", color = "grey", size = 0.5) +
geom_errorbar(aes(ymin = logchange-1.96*sd_logchange,
ymax =logchange + 1.96*sd_logchange, linetype = Line_type),
width = 0.2, size = 0.5, alpha = 0.6,position = pd) +
geom_point(position = pd, fill = "white", size =3) +
ylab("Change in egg-to-adult viability\nbetween intermediate and initial phenotyping steps") +
xlab("Populations") +
labs(shape = "Test fruit", color = "Selection fruit") +
guides(color = FALSE,
shape = guide_legend(override.aes = list(fill = c("black")))) +
scale_shape_manual(values = c(21, 22, 24)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
scale_fill_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
theme_LO_sober + theme(axis.text.x = element_blank()) +
coord_cartesian(expand = FALSE, ylim = c(-2, 2), xlim=c(0, 15),clip = "off") +
ggtitle("Intermediate phenotyping step") +
annotate("text", x = 2.5, y =1.5, label = 'bold(" Evolved\non cherry")',
size =3,colour = "#BC3C6D",parse = TRUE) +
annotate("text", x = 7.5, y =1.5, label = 'bold(" Evolved\non cranberry")',
size =3,colour = "#FDB424",parse = TRUE) +
annotate("text", x = 12.5, y =1.5, label = 'bold(" Evolved\non strawberry")',
size =3,colour = "#3FAA96",parse = TRUE)
symp_allop_g7
################### FINAL PHENOTYPING
# Re-order levels of Line
data_sum_G29 <- data_logchangeEmergence_rate[data_logchangeEmergence_rate$Generation == "29",]
data_sum_G29$Line <- factor(data_sum_G29$Line, levels = c("CEA", "CEB", "CEC",
"CRD", "CRA", "CRC", "CRB", "CRE",
"FRA", "FRC", "FRB"))
symp_allop_G29 <- ggplot(data = data_sum_G29,
aes(x = Line, y=logchange, group = Treatment,
color =Fruit_s, shape = Treatment,fill =Fruit_s)) +
geom_hline(yintercept = 0, linetype = "dashed", color = "grey", size = 0.5) +
geom_point(size =3, position = pd) +
geom_errorbar(aes(ymin =logchange-1.96*sd_logchange, ymax =logchange + 1.96*sd_logchange),
width= 0.2, size = 0.5, alpha = 0.6,position = pd) +
ylab("Change in egg-to-adult viability\nbetween final and initial phenotyping steps") +
xlab("Populations") +
labs(shape = "Test fruit", color = "Selection fruit") +
guides(fill = FALSE, alpha = FALSE) +
scale_shape_manual(values = c(21, 22, 24)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
scale_fill_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
theme_LO_sober + theme(axis.text.x = element_blank()) +
coord_cartesian(expand = FALSE, ylim = c(-2, 2), xlim=c(0, 11),clip = "off") +
annotate("text", x = 2, y =1.5, label = 'bold(" Evolved\non cherry")',
size =3,colour = "#BC3C6D",parse = TRUE) +
annotate("text", x = 6, y =1.5, label = 'bold(" Evolved\non cranberry")',
size =3,colour = "#FDB424",parse = TRUE) +
annotate("text", x = 10, y =1.5, label = 'bold(" Evolved\non strawberry")',
size =3,colour = "#3FAA96",parse = TRUE) +
ggtitle("Final phenotyping step")
symp_allop_G29
### Add stroke
DIF_FITNESS_G7 <- symp_allop_g7 +
geom_point(aes(alpha = SA, color = interaction(SA,Fruit_s)), position = pd, size =3, stroke =1.5, fill = "white") +
scale_alpha_manual(values = c(0, 1)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96", "#7C2748", "#CA8702", "#328677", "#BC3C6D", "#FDB424", "#3FAA96"))
DIF_FITNESS_G7
DIF_FITNESS_G29 <- symp_allop_G29 + geom_point(aes(alpha = SA, color = interaction(SA,Fruit_s)), position = pd, size =3, stroke =1.5) +
scale_alpha_manual(values = c(0, 1)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96", "#7C2748", "#CA8702", "#328677", "#BC3C6D", "#FDB424", "#3FAA96"))
DIF_FITNESS_G29
plot_legend <- ggplot(data = data_logchange, aes(x = Line, y=logchange, group = Treatment,
color =Fruit_s, shape = Treatment,fill =Fruit_s)) +
geom_point(size =3, position = pd) +
labs(shape = "Test fruit", color = "Selection fruit") +
scale_shape_manual(values = c(21, 22, 24)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
theme_LO_sober +
guides(fill = FALSE) +
theme(legend.key.size = unit(0.5, "cm"),
axis.text.x = element_blank()) +
coord_cartesian(expand = FALSE, ylim = c(-2, 2), xlim=c(0, 11),clip = "off")
###########################################################
#############################################ALL PLOT
legend_trade <- lemon::g_legend(plot_legend)
SYMP_ALLOP_TOTAL <- cowplot::ggdraw() +
cowplot::draw_plot(DIF_FITNESS_G7 + theme(legend.position = 'none',
plot.margin =unit(c(0.5, 3.5, 0.5, 0.5), "cm")),
x = 0, y = 0.5, width = 1, height = 0.5) +
cowplot::draw_plot(DIF_FITNESS_G29 + theme(legend.position = 'none',
plot.margin =unit(c(0.5, 3.5, 0.5, 0.5), "cm")),
x = 0, y = 0, width = 1, height = 0.5) +
cowplot::draw_plot_label(c("A", "B"),
x = c(0, 0), y = c(1, 0.5),
size = 14) +
cowplot::draw_plot(legend_trade, x = 0.93, y = 0.5, width = 0.001, height = 0.001)
SYMP_ALLOP_TOTAL
cowplot::save_plot(file =here::here("figures", "FIG_SX_HeterogeneityEgg-to-adultviability.pdf"),
SYMP_ALLOP_TOTAL, base_height = 20/cm(1), base_width = 20/cm(1), dpi = 1200)
SYMP_ALLOP_TOTAL <- cowplot::ggdraw() +
cowplot::draw_plot(DIF_FITNESS_G7Nb_eggs + theme(legend.position = 'none',
plot.margin =unit(c(0.5, 3.5, 0.5, 0.5), "cm")),
x = 0, y = 0.5, width = 1, height = 0.5) +
cowplot::draw_plot(DIF_FITNESS_G29Nb_eggs + theme(legend.position = 'none',
plot.margin =unit(c(0.5, 3.5, 0.5, 0.5), "cm")),
x = 0, y = 0, width = 1, height = 0.5) +
cowplot::draw_plot(DIF_FITNESS_G7 + theme(legend.position = 'none',
plot.margin =unit(c(0.5, 3.5, 0.5, 0.5), "cm")),
x = 0.5, y = 0.5, width = 1, height = 0.5) +
cowplot::draw_plot(DIF_FITNESS_G29 + theme(legend.position = 'none',
plot.margin =unit(c(0.5, 3.5, 0.5, 0.5), "cm")),
x = 0.5, y = 0, width = 1, height = 0.5) +
cowplot::draw_plot_label(c("A", "B", "C", "D"),
x = c(0, 0, 0.5, 0.5), y = c(1, 0.5, 1, 0.5),
size = 14) +
cowplot::draw_plot(legend_trade, x = 0.93, y = 0.5, width = 0.001, height = 0.001)
cowplot::save_plot(file =here::here("figures", "FIG_SX_HeterogeneityNb_Egges_Egg-to-adultviability.pdf"),
SYMP_ALLOP_TOTAL, base_height = 20/cm(1), base_width = 30/cm(1), dpi = 1200)5 CORRELATED RESPONSES
5.1 Analysis MA and correlation
# ##### Create empty dataframe with slopes and confidence interval for all pairwise
Estimates_pairwise <- data.frame(Variables = rep(rep(c("correlation","cor_CI_inf","cor_CI_sup",
"slope", "intercept"),2),3),
Generation = rep(c(rep(7, 5), rep(29, 5)),3),
Pairwise = rep(c("Cherry_Cranberry",
"Cranberry_Strawberry",
"Strawberry_Cherry"), each=10),
Estimates = NA)
#########################
######################### G7
#########################
#######
####### Cherry Cranberry
#######
## Compute weighted correlation
weightedcor <- sjstats:::weighted_correlation(TEMP_dataG7_CheCran, x=logchange_symp, y=logchange_allop, weights=N_sumsympallop, ci.lvl = 0.95)
## Compute Major Axis regression
model <- lmodel2::lmodel2(logchange_allop ~ logchange_symp,
range.y = "interval",range.x = "interval",
data = TEMP_dataG7_CheCran[rep(1:nrow(TEMP_dataG7_CheCran), TEMP_dataG7_CheCran$N_sumsympallop),], nperm=0)
## The estimation of the confidence interval with cor.test is wrong
cor.test(x=TEMP_dataG7_CheCran[rep(1:nrow(TEMP_dataG7_CheCran), TEMP_dataG7_CheCran$N_sumsympallop), "logchange_symp"], y=TEMP_dataG7_CheCran[rep(1:nrow(TEMP_dataG7_CheCran), TEMP_dataG7_CheCran$N_sumsympallop), "logchange_allop"])##
## Pearson's product-moment correlation
##
## data: TEMP_dataG7_CheCran[rep(1:nrow(TEMP_dataG7_CheCran), TEMP_dataG7_CheCran$N_sumsympallop), and TEMP_dataG7_CheCran[rep(1:nrow(TEMP_dataG7_CheCran), TEMP_dataG7_CheCran$N_sumsympallop), "logchange_symp"] and "logchange_allop"]
## t = 6.3642, df = 136, p-value = 2.79e-09
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.3390707 0.5982491
## sample estimates:
## cor
## 0.4790334# Fill in data
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="correlation"] <- weightedcor$estimate
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="cor_CI_inf"] <- weightedcor$ci[1]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="cor_CI_sup"] <- weightedcor$ci[2]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="intercept"] <- model$regression.results[2, 2]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="slope"] <- model$regression.results[2, 3]
#######
####### Cranberry Strawberry
#######
## Compute weighted correlation
weightedcor <- sjstats:::weighted_correlation(TEMP_dataG7_CranStraw, x=logchange_symp, y=logchange_allop, weights=N_sumsympallop, ci.lvl = 0.95)
## Compute Major Axis regression
model <- lmodel2::lmodel2(logchange_allop ~ logchange_symp,
range.y = "interval",range.x = "interval",
data = TEMP_dataG7_CranStraw[rep(1:nrow(TEMP_dataG7_CranStraw), TEMP_dataG7_CranStraw$N_sumsympallop),], nperm=0)
# Fill in data
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="correlation"] <- weightedcor$estimate
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="cor_CI_inf"] <- weightedcor$ci[1]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="cor_CI_sup"] <- weightedcor$ci[2]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="intercept"] <- model$regression.results[2, 2]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="slope"] <- model$regression.results[2, 3]
#######
####### Strawberry Cherry
#######
## Compute weighted correlation
weightedcor <- sjstats:::weighted_correlation(TEMP_dataG7_StrawChe, x=logchange_symp, y=logchange_allop, weights=N_sumsympallop, ci.lvl = 0.95)
## Compute Major Axis regression
model <- lmodel2::lmodel2(logchange_allop ~ logchange_symp,
range.y = "interval",range.x = "interval",
data = TEMP_dataG7_StrawChe[rep(1:nrow(TEMP_dataG7_StrawChe), TEMP_dataG7_StrawChe$N_sumsympallop),], nperm=0)
# Fill in data
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="correlation"] <- weightedcor$estimate
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="cor_CI_inf"] <- weightedcor$ci[1]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="cor_CI_sup"] <- weightedcor$ci[2]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="intercept"] <- model$regression.results[2, 2]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="slope"] <- model$regression.results[2, 3]
#########################
######################### G29
#########################
#######
####### Cherry Cranberry
#######
## Compute unweighted correlation
unweightedcor <- cor.test(x=TEMP_dataG29_CheCran$logchange_symp, y=TEMP_dataG29_CheCran$logchange_allop)
## Compute weighted correlation
weightedcor <- sjstats:::weighted_correlation(TEMP_dataG29_CheCran, x=logchange_symp, y=logchange_allop, weights=N_sumsympallop, ci.lvl = 0.95)
## Compute Major Axis regression
model <- lmodel2::lmodel2(logchange_allop ~ logchange_symp,
range.y = "interval",range.x = "interval",
data = TEMP_dataG29_CheCran[rep(1:nrow(TEMP_dataG29_CheCran), TEMP_dataG29_CheCran$N_sumsympallop),], nperm=0)
# Fill in data
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="correlation"] <- unweightedcor$estimate
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="cor_CI_inf"] <- unweightedcor$conf.int[1]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="cor_CI_sup"] <- unweightedcor$conf.int[2]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="intercept"] <- model$regression.results[2, 2]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="slope"] <- model$regression.results[2, 3]
#######
####### Cranberry Strawberry
#######
## Compute unweighted correlation
unweightedcor <- cor.test(x=TEMP_dataG29_CranStraw$logchange_symp, y=TEMP_dataG29_CranStraw$logchange_allop)
## Compute weighted correlation
weightedcor <- sjstats:::weighted_correlation(TEMP_dataG29_CranStraw, x=logchange_symp, y=logchange_allop, weights=N_sumsympallop, ci.lvl = 0.95)
## Compute Major Axis regression
model <- lmodel2::lmodel2(logchange_allop ~ logchange_symp,
range.y = "interval",range.x = "interval",
data = TEMP_dataG29_CranStraw[rep(1:nrow(TEMP_dataG29_CranStraw), TEMP_dataG29_CranStraw$N_sumsympallop),], nperm=0)
# Fill in data
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="correlation"] <- unweightedcor$estimate
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="cor_CI_inf"] <- unweightedcor$conf.int[1]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="cor_CI_sup"] <- unweightedcor$conf.int[2]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="intercept"] <- model$regression.results[2, 2]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="slope"] <- model$regression.results[2, 3]
#######
####### Strawberry Cherry
#######
## Compute unweighted correlation
unweightedcor <- cor.test(x=TEMP_dataG29_StrawChe$logchange_symp, y=TEMP_dataG29_StrawChe$logchange_allop)
## Compute weighted correlation
weightedcor <- sjstats:::weighted_correlation(TEMP_dataG29_StrawChe, x=logchange_symp, y=logchange_allop, weights=N_sumsympallop, ci.lvl = 0.95)
## Compute Major Axis regression
model <- lmodel2::lmodel2(logchange_allop ~ logchange_symp,
range.y = "interval",range.x = "interval",
data = TEMP_dataG29_StrawChe[rep(1:nrow(TEMP_dataG29_StrawChe), TEMP_dataG29_StrawChe$N_sumsympallop),], nperm=0)
# Fill in data
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="correlation"] <- unweightedcor$estimate
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="cor_CI_inf"] <- unweightedcor$conf.int[1]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="cor_CI_sup"] <- unweightedcor$conf.int[2]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="intercept"] <- model$regression.results[2, 2]
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="slope"] <- model$regression.results[2, 3]
Estimates_pairwise## Variables Generation Pairwise Estimates
## 1 correlation 7 Cherry_Cranberry 0.479033353
## 2 cor_CI_inf 7 Cherry_Cranberry 0.001431407
## 3 cor_CI_sup 7 Cherry_Cranberry 0.956635299
## 4 slope 7 Cherry_Cranberry 4.160858972
## 5 intercept 7 Cherry_Cranberry -0.356412153
## 6 correlation 29 Cherry_Cranberry -0.567557085
## 7 cor_CI_inf 29 Cherry_Cranberry -0.908773300
## 8 cor_CI_sup 29 Cherry_Cranberry 0.228504360
## 9 slope 29 Cherry_Cranberry -1.427127889
## 10 intercept 29 Cherry_Cranberry 0.268405728
## 11 correlation 7 Cranberry_Strawberry 0.535360085
## 12 cor_CI_inf 7 Cranberry_Strawberry 0.105487696
## 13 cor_CI_sup 7 Cranberry_Strawberry 0.965232474
## 14 slope 7 Cranberry_Strawberry 1.070761415
## 15 intercept 7 Cranberry_Strawberry -0.049525449
## 16 correlation 29 Cranberry_Strawberry 0.593494172
## 17 cor_CI_inf 29 Cranberry_Strawberry -0.191100790
## 18 cor_CI_sup 29 Cranberry_Strawberry 0.915350061
## 19 slope 29 Cranberry_Strawberry 1.613908561
## 20 intercept 29 Cranberry_Strawberry -0.582936853
## 21 correlation 7 Strawberry_Cherry 0.616674051
## 22 cor_CI_inf 7 Strawberry_Cherry 0.188354905
## 23 cor_CI_sup 7 Strawberry_Cherry 1.044993196
## 24 slope 7 Strawberry_Cherry 0.829196290
## 25 intercept 7 Strawberry_Cherry 0.177499018
## 26 correlation 29 Strawberry_Cherry 0.676000585
## 27 cor_CI_inf 29 Strawberry_Cherry -0.300322260
## 28 cor_CI_sup 29 Strawberry_Cherry 0.960575095
## 29 slope 29 Strawberry_Cherry 2.537639762
## 30 intercept 29 Strawberry_Cherry -0.7647082205.2 Compute confidence interval of the difference
computeCIcordifG29G7(pair="Cherry_Cranberry")## dif lower higher
## -1.0465904 -1.9126980 -0.1804829computeCIcordifG29G7(pair="Cranberry_Strawberry")## dif lower higher
## 0.05813409 -0.78991100 0.90617917computeCIcordifG29G7(pair="Strawberry_Cherry")## dif lower higher
## 0.05932653 -0.95762427 1.07627734# ##### Create empty dataframe with slopes and confidence interval for all pairwise
Estimates_pairwise <- data.frame(Variables = rep(rep(c("correlation","cor_CI_inf","cor_CI_sup",
"slope", "intercept"),2),3),
Generation = rep(c(rep(7, 5), rep(29, 5)),3),
Pairwise = rep(c("Cherry_Cranberry",
"Cranberry_Strawberry",
"Strawberry_Cherry"), each=10),
Estimates = NA ,
EstimatesWeighed = NA ,
EstimatesUnweighed = NA )
# Calculate sd in case of NA value
mean_sd_G7 <- sqrt(mean(data_logchange$sd_logchange[data_logchange$Generation=="7"]^2, na.rm = TRUE))
mean_sd_G29 <- sqrt(mean(data_logchange$sd_logchange[data_logchange$Generation=="29"]^2, na.rm = TRUE))
## Check whether the squared standard errors are of the same magnitude across fruit media (answer:yes)
tapply(data_logchange$sd_logchange^2, list(data_logchange$Treatment, data_logchange$Generation), mean, na.rm = TRUE)
## Impute directly the missing standard errors
data_logchange$se_logchange <- data_logchange$sd_logchange
data_logchange$se_logchange[data_logchange$Generation=="7"&is.na(data_logchange$se_logchange)] <- rep(mean_sd_G7, sum(data_logchange$Generation=="7"&is.na(data_logchange$se_logchange)))
## Nbr sample
nb_sim = 5000
#########################
######################### G7
#########################
#######
####### Cherry Cranberry
#######
# MA axis regression with bootstrap to estimate slope, intercept and
sim <- data.frame(t(sapply(1:nb_sim, bootstrap_lmodel2, data_pairwise = TEMP_dataG7_CheCran, sd_NA = mean_sd_G7)))
# Fill out data with quantile of simulations
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="correlation"] <- mean(sim$correlation.cor, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="cor_CI_inf"] <- quantile(sim$correlation.cor, probs = 0.025, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="cor_CI_sup"] <- quantile(sim$correlation.cor, probs = 0.975, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="intercept"] <- mean(sim$intercept, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="slope"] <- mean(sim$slope, na.rm = TRUE)
## Add other correlations
Estimates_pairwise$EstimatesWeighed[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="correlation"] <- weighted_correlation(TEMP_dataG7_CheCran, x=logchange_symp, y=logchange_allop, weights=N_sumsympallop, ci.lvl = 0.95)$estimate
Estimates_pairwise$EstimatesWeighed[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="cor_CI_inf"] <- weighted_correlation(TEMP_dataG7_CheCran, x=logchange_symp, y=logchange_allop, weights=N_sumsympallop, ci.lvl = 0.95)$ci[1]
Estimates_pairwise$EstimatesWeighed[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="cor_CI_sup"] <- weighted_correlation(TEMP_dataG7_CheCran, x=logchange_symp, y=logchange_allop, weights=N_sumsympallop, ci.lvl = 0.95)$ci[2]
Estimates_pairwise$EstimatesUnweighed[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="correlation"] <- cor.test(x=TEMP_dataG7_CheCran$logchange_symp, y=TEMP_dataG7_CheCran$logchange_allop)$estimate
Estimates_pairwise$EstimatesUnweighed[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="cor_CI_inf"] <- cor.test(x=TEMP_dataG7_CheCran$logchange_symp, y=TEMP_dataG7_CheCran$logchange_allop)$conf.int[1]
Estimates_pairwise$EstimatesUnweighed[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="cor_CI_sup"] <- cor.test(x=TEMP_dataG7_CheCran$logchange_symp, y=TEMP_dataG7_CheCran$logchange_allop)$conf.int[2]
#Save sim G7 to calculate proportion later:
sim_CheCranG7<-sim
#######
####### Cranberry Strawberry
#######
# MA axis regression with bootstrap to estimate slope, intercept and
sim <- data.frame(t(sapply(1:nb_sim, bootstrap_lmodel2, data_pairwise = TEMP_dataG7_CranStraw, sd_NA = mean_sd_G7)))
# Fill out data with quantile of simulations
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="correlation"] <- quantile(sim$cor, probs = 0.5, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="cor_CI_inf"] <- quantile(sim$cor, probs = 0.025, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="cor_CI_sup"] <- quantile(sim$cor, probs = 0.975, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="intercept"] <- quantile(sim$intercept,probs = 0.5, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="slope"] <- quantile(sim$slope,probs = 0.5, na.rm = TRUE)
## Save sim G7 to calculate correlation difference later:
sim_CranStrawG7<-sim
#######
####### Strawberry Cherry
#######
# MA axis regression with bootstrap to estimate slope, intercept and
sim <- data.frame(t(sapply(1:nb_sim, bootstrap_lmodel2, data_pairwise = TEMP_dataG7_StrawChe, sd_NA = mean_sd_G7)))
# Fill out data with quantile of simulations
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="correlation"] <- quantile(sim$cor, probs = 0.5, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="cor_CI_inf"] <- quantile(sim$cor, probs = 0.025, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="cor_CI_sup"] <- quantile(sim$cor, probs = 0.975, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="intercept"] <- quantile(sim$intercept,probs = 0.5, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==7&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="slope"] <- quantile(sim$slope,probs = 0.5, na.rm = TRUE)
## Save sim G7 to calculate correlation difference later:
sim_StrawCheG7<-sim
#########################
######################### G29
#########################
#######
####### Cherry Cranberry
#######
# MA axis regression with bootstrap to estimate slope, intercept and
sim <- data.frame(t(sapply(1:nb_sim, bootstrap_lmodel2, data_pairwise = TEMP_dataG29_CheCran, sd_NA = mean_sd_G29)))
# Fill out data with quantile of simulations
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="correlation"] <- quantile(sim$cor, probs = 0.5, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="cor_CI_inf"] <- quantile(sim$cor, probs = 0.025, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="cor_CI_sup"] <- quantile(sim$cor, probs = 0.975, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="intercept"] <- quantile(sim$intercept, probs = 0.5, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="slope"] <- quantile(sim$slope, probs = 0.5, na.rm = TRUE)
## Save sim G29 to calculate correlation difference later:
sim_CheCranG29<-sim
#######
####### Cranberry Strawberry
#######
# MA axis regression with bootstrap to estimate slope, intercept and
sim <- data.frame(t(sapply(1:nb_sim, bootstrap_lmodel2, data_pairwise = TEMP_dataG29_CranStraw, sd_NA = mean_sd_G29)))
# Fill out data with quantile of simulations
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="correlation"] <- quantile(sim$cor, probs = 0.5, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="cor_CI_inf"] <- quantile(sim$cor, probs = 0.025, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="cor_CI_sup"] <- quantile(sim$cor, probs = 0.975, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="intercept"] <- quantile(sim$intercept,probs = 0.5, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cranberry_Strawberry"&
Estimates_pairwise$Variables=="slope"] <- quantile(sim$slope,probs = 0.5, na.rm = TRUE)
## Save sim G29 to calculate correlation difference later:
sim_CranStrawG29<-sim
#######
####### Strawberry Cherry
#######
rm(sim)
# MA axis regression with bootstrap to estimate slope, intercept and
sim <- data.frame(t(sapply(1:nb_sim, bootstrap_lmodel2, data_pairwise = TEMP_dataG29_StrawChe, sd_NA = mean_sd_G29)))
# Fill out data with quantile of simulations
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="correlation"] <- quantile(sim$cor, probs = 0.5, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="cor_CI_inf"] <- quantile(sim$cor, probs = 0.025, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="cor_CI_sup"] <- quantile(sim$cor, probs = 0.975, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="intercept"] <- quantile(sim$intercept,probs = 0.5, na.rm = TRUE)
Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Strawberry_Cherry"&
Estimates_pairwise$Variables=="slope"] <- quantile(sim$slope,probs = 0.5, na.rm = TRUE)
Estimates_pairwise
## Save sim G29 to calculate correlation difference later:
sim_StrawCheG29<-sim
## Compute the confidence interval of the difference between correlations in G7 and G29
difcorCheCran <- sim_CheCranG29$correlation.cor-sim_CheCranG7$correlation.cor
quantile(difcorCheCran, probs = c(0.025, 0.975), na.rm=TRUE)
difcorCranStraw <- sim_CranStrawG29$correlation.cor-sim_CranStrawG7$correlation.cor
quantile(difcorCranStraw, probs = c(0.025, 0.975), na.rm=TRUE)
difcorStrawChe <- sim_StrawCheG29$correlation.cor-sim_StrawCheG7$correlation.cor
quantile(difcorStrawChe, probs = c(0.025, 0.975), na.rm=TRUE)
#Cherry Cranbery
#Compute proportion of simulations correlations G7 are lower than correlation estimated on G29
cor_G29_CheCran<-Estimates_pairwise$Estimates[Estimates_pairwise$Generation==29&
Estimates_pairwise$Pairwise=="Cherry_Cranberry"&
Estimates_pairwise$Variables=="correlation"]
#Estimates of correlation during final phenotyping for Cherry Cranberry
p_val_decreaseCheCran<-length(sim_CheCranG7$correlation.cor[sim_CheCranG7$correlation.cor<cor_G29_CheCran])/nb_sim
p_val_decreaseCheCran5.3 Plot
ymin = -50
ymax = 50
## Function equation R
eq_r <- function(gen = 7, pair = "Cherry_Cranberry") {
if(gen == 7 &pair == "Cherry_Cranberry"){
as.character(
as.expression(
substitute(~~italic(r)^2~"="~r2~"["~inf~","~sup~"]",
list(r2 = format(Estimates_pairwise$Estimates[Estimates_pairwise$Generation == gen &
Estimates_pairwise$Pairwise == pair &
Estimates_pairwise$Variables == "correlation"], digits = 2),
inf = format(Estimates_pairwise$Estimates[Estimates_pairwise$Generation == gen &
Estimates_pairwise$Pairwise == pair &
Estimates_pairwise$Variables == "cor_CI_inf"], digits = 1, nsmall=2),
sup = format(Estimates_pairwise$Estimates[Estimates_pairwise$Generation == gen &
Estimates_pairwise$Pairwise == pair &
Estimates_pairwise$Variables == "cor_CI_sup"], digits = 2)))
)
)
}else{
as.character(
as.expression(
substitute(~~italic(r)^2~"="~r2~"["~inf~","~sup~"]",
list(r2 = format(Estimates_pairwise$Estimates[Estimates_pairwise$Generation == gen &
Estimates_pairwise$Pairwise == pair &
Estimates_pairwise$Variables == "correlation"], digits = 2),
inf = format(Estimates_pairwise$Estimates[Estimates_pairwise$Generation == gen &
Estimates_pairwise$Pairwise == pair &
Estimates_pairwise$Variables == "cor_CI_inf"], digits = 2, nsmall=2),
sup = format(Estimates_pairwise$Estimates[Estimates_pairwise$Generation == gen &
Estimates_pairwise$Pairwise == pair &
Estimates_pairwise$Variables == "cor_CI_sup"], digits = 2)))
)
)
}
}
##################################################
################## G7
# Limits
ymin_CheCranG7=min(min(TEMP_dataG7_CheCran$logchange_allop-1.96*TEMP_dataG7_CheCran$sd_allop, na.rm= TRUE),
min(TEMP_dataG7_CheCran$logchange_symp-1.96*TEMP_dataG7_CheCran$sd_symp, na.rm= TRUE))
ymax_CheCranG7=max(max(TEMP_dataG7_CheCran$logchange_allop + 1.96*TEMP_dataG7_CheCran$sd_allop, na.rm= TRUE),
max(TEMP_dataG7_CheCran$logchange_symp + 1.96*TEMP_dataG7_CheCran$sd_symp, na.rm= TRUE))
lim_text<-ymin_CheCranG7+0.99*(ymax_CheCranG7-ymin_CheCranG7)
# Plot
CheCran_G7 <- ggplot(data = TEMP_dataG7_CheCran) +
geom_abline(intercept = Estimates_pairwise$Estimates[Estimates_pairwise$Generation == 7 &
Estimates_pairwise$Pairwise == "Cherry_Cranberry" &
Estimates_pairwise$Variables == "intercept"],
slope = Estimates_pairwise$Estimates[Estimates_pairwise$Generation == 7 &
Estimates_pairwise$Pairwise == "Cherry_Cranberry" &
Estimates_pairwise$Variables == "slope"],
colour = "black", size = 0.75) +
geom_text(x = -0.825, y = lim_text, label = eq_r(gen = 7, pair = "Cherry_Cranberry"), parse = TRUE, color="black", size = 3.5) +
geom_errorbar(aes(x = logchange_symp, ymin = logchange_allop-(1.96*sd_allop),
ymax = logchange_allop + (1.96*sd_allop),
color = Symp, linetype = Line_type),
width= 0.02, size = 0.3, alpha = 0.8) +
geom_errorbarh(aes(y = logchange_allop, xmin = logchange_symp-1.96*sd_symp, xmax = logchange_symp + 1.96*sd_symp,
color = Symp, linetype = Line_type),
height = 0.02, size = 0.3, alpha = 0.8) +
geom_point(aes(x = logchange_symp, y = logchange_allop, color = Symp,fill = Symp, shape = Allop),
size = 3, fill = "white", stroke =1.2) +
xlab("Fitness change in\nselective environment") +
ylab("Fitness change in\nalternative environment") +
ggtitle("Cherry vs. Cranberry") +
theme_LO_sober +
labs(shape = "Test fruit", color = "Evolution fruit") +
scale_shape_manual(values = c(21, 22)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424")) +
coord_cartesian(ylim = c(ymin_CheCranG7, ymax_CheCranG7),
xlim = c(ymin_CheCranG7, ymax_CheCranG7))
CheCran_G7
# Limits
ymin_CranStrawG7=min(min(TEMP_dataG7_CranStraw$logchange_allop-1.96*TEMP_dataG7_CranStraw$sd_allop, na.rm= TRUE),
min(TEMP_dataG7_CranStraw$logchange_symp-1.96*TEMP_dataG7_CranStraw$sd_symp, na.rm= TRUE))
ymax_CranStrawG7=max(max(TEMP_dataG7_CranStraw$logchange_allop + 1.96*TEMP_dataG7_CranStraw$sd_allop, na.rm= TRUE),
max(TEMP_dataG7_CranStraw$logchange_symp + 1.96*TEMP_dataG7_CranStraw$sd_symp, na.rm= TRUE))
lim_text<-ymin_CranStrawG7+0.99*(ymax_CranStrawG7-ymin_CranStrawG7)
CranStraw_G7 <- ggplot(data = TEMP_dataG7_CranStraw) +
geom_abline(intercept = Estimates_pairwise$Estimates[Estimates_pairwise$Generation == 7 &
Estimates_pairwise$Pairwise == "Cranberry_Strawberry" &
Estimates_pairwise$Variables == "intercept"],
slope = Estimates_pairwise$Estimates[Estimates_pairwise$Generation == 7 &
Estimates_pairwise$Pairwise == "Cranberry_Strawberry" &
Estimates_pairwise$Variables == "slope"],
colour = "black", size = 0.75) +
geom_errorbar(aes(x =logchange_symp, ymin = logchange_allop-(1.96*sd_allop),
ymax = logchange_allop + (1.96*sd_allop),
color = Symp, linetype = Line_type),
width= 0.02, size = 0.3, alpha = 0.8) +
geom_errorbarh(aes(y = logchange_allop, xmin = logchange_symp-1.96*sd_symp, xmax = logchange_symp + 1.96*sd_symp,
color = Symp, linetype = Line_type),
height = 0.02, size = 0.3, alpha = 0.8) +
geom_point(aes(x =logchange_symp, y = logchange_allop, color = Symp,fill = Symp, shape = Allop),
size =3, fill = "white", stroke =1.2) +
geom_text(x = -0.45, y = lim_text, label = eq_r(gen = 7, pair = "Cranberry_Strawberry"), parse = TRUE, color="black", size = 3.5) +
xlab("Fitness change in\nselective environment") +
ylab("Fitness change in\nalternative environment") +
ggtitle("Cranberry vs. Strawberry") +
coord_cartesian(ylim = c(ymin_CranStrawG7, ymax_CranStrawG7),
xlim = c(ymin_CranStrawG7, ymax_CranStrawG7)) +
labs(shape = "Test fruit", color = "Selection fruit") +
scale_shape_manual(values = c(22, 24)) +
scale_color_manual(values = c("#FDB424", "#3FAA96")) +
theme_LO_sober
CranStraw_G7
# Limits
ymin_StrawCheG7=min(min(TEMP_dataG7_StrawChe$logchange_allop-1.96*TEMP_dataG7_StrawChe$sd_allop, na.rm= TRUE),
min(TEMP_dataG7_StrawChe$logchange_symp-1.96*TEMP_dataG7_StrawChe$sd_symp, na.rm= TRUE))
ymax_StrawCheG7=max(max(TEMP_dataG7_StrawChe$logchange_allop + 1.96*TEMP_dataG7_StrawChe$sd_allop, na.rm= TRUE),
max(TEMP_dataG7_StrawChe$logchange_symp + 1.96*TEMP_dataG7_StrawChe$sd_symp, na.rm= TRUE))
lim_text<-ymin_StrawCheG7+0.99*(ymax_StrawCheG7-ymin_StrawCheG7)
StrawChe_G7 <- ggplot(data = TEMP_dataG7_StrawChe) +
geom_abline(intercept = Estimates_pairwise$Estimates[Estimates_pairwise$Generation == 7 &
Estimates_pairwise$Pairwise == "Strawberry_Cherry" &
Estimates_pairwise$Variables == "intercept"],
slope = Estimates_pairwise$Estimates[Estimates_pairwise$Generation == 7 &
Estimates_pairwise$Pairwise == "Strawberry_Cherry" &
Estimates_pairwise$Variables == "slope"],
colour = "black", size = 0.75) +
geom_errorbar(aes(x =logchange_symp, ymin = logchange_allop-(1.96*sd_allop), ymax = logchange_allop + (1.96*sd_allop),
color = Symp, linetype = Line_type),
width= 0.02, size = 0.3, alpha = 0.8) +
geom_errorbarh(aes(y = logchange_allop, xmin = logchange_symp-1.96*sd_symp, xmax = logchange_symp + 1.96*sd_symp,
color = Symp, linetype = Line_type),
height = 0.02, size = 0.3, alpha = 0.8) +
geom_point(aes(x = logchange_symp, y = logchange_allop, color = Symp,fill = Symp, shape = Allop),
size =3, fill = "white", stroke =1.2) +
geom_text(x = -0.83, y = lim_text, label = eq_r(gen = 7, pair = "Strawberry_Cherry"), parse = TRUE, color="black", size = 3.5) +
xlab("Fitness change in\nselective environment") +
ylab("Fitness change in\nalternative environment") +
ggtitle("Strawberry vs. Cherry") +
labs(shape = "Test fruit", color = "Selection fruit") +
scale_shape_manual(values = c(21, 24)) +
scale_color_manual(values = c("#BC3C6D", "#3FAA96")) +
coord_cartesian(ylim = c(ymin_StrawCheG7, ymax_StrawCheG7),
xlim = c(ymin_StrawCheG7, ymax_StrawCheG7)) +
theme_LO_sober
StrawChe_G7
#####################################################################################
################################## G29 ##################################
#####################################################################################
TEMP_dataG29_CheCran## Line Treatment Fruit_s Generation N_allop logchange sd_logchange SA
## 15 CEA Cranberry Cherry 29 30 -0.21774071 0.13281291 0
## 17 CEB Cranberry Cherry 29 30 -0.59200204 0.09634940 0
## 21 CEC Cranberry Cherry 29 30 -0.20670829 0.12591289 0
## 37 CRA Cherry Cranberry 29 30 0.32537794 0.10025669 0
## 41 CRB Cherry Cranberry 29 30 0.16929862 0.12699030 0
## 45 CRC Cherry Cranberry 29 30 0.08474123 0.12334598 0
## 47 CRD Cherry Cranberry 29 30 0.54736475 0.09718018 0
## 49 CRE Cherry Cranberry 29 30 -0.10496044 0.11556266 0
## Symp Allop Line_type logchange_allop sd_allop logchange_symp
## 15 Cherry Cranberry dashed -0.21774071 0.13281291 0.14767000
## 17 Cherry Cranberry dashed -0.59200204 0.09634940 0.25038029
## 21 Cherry Cranberry dashed -0.20670829 0.12591289 0.35854062
## 37 Cranberry Cherry dashed 0.32537794 0.10025669 -0.14885350
## 41 Cranberry Cherry dashed 0.16929862 0.12699030 0.29719689
## 45 Cranberry Cherry dashed 0.08474123 0.12334598 0.02703717
## 47 Cranberry Cherry dashed 0.54736475 0.09718018 -0.15627744
## 49 Cranberry Cherry dashed -0.10496044 0.11556266 0.72513486
## sd_symp N_symp N_sumsympallop
## 15 0.1574284 30 60
## 17 0.1009864 30 60
## 21 0.1169300 30 60
## 37 0.1193552 30 60
## 41 0.1531197 30 60
## 45 0.1156381 30 60
## 47 0.1156695 30 60
## 49 0.1275430 30 60TEMP_dataG29_CranStraw## Line Treatment Fruit_s Generation N_allop logchange sd_logchange SA
## 39 CRA Strawberry Cranberry 29 30 -0.29731422 0.1098141 0
## 42 CRB Strawberry Cranberry 29 30 -0.24539416 0.1387043 0
## 43 CRC Strawberry Cranberry 29 30 -0.28950778 0.2012393 0
## 46 CRD Strawberry Cranberry 29 30 -1.02555272 0.1902082 0
## 51 CRE Strawberry Cranberry 29 30 -0.16810249 0.1398057 0
## 67 FRA Cranberry Strawberry 29 30 -0.25158762 0.1712755 0
## 72 FRB Cranberry Strawberry 29 30 0.04303751 0.1441119 0
## 73 FRC Cranberry Strawberry 29 30 0.52109773 0.1254820 0
## Symp Allop Line_type logchange_allop sd_allop logchange_symp
## 39 Cranberry Strawberry dashed -0.29731422 0.1098141 -0.14885350
## 42 Cranberry Strawberry dashed -0.24539416 0.1387043 0.29719689
## 43 Cranberry Strawberry dashed -0.28950778 0.2012393 0.02703717
## 46 Cranberry Strawberry dashed -1.02555272 0.1902082 -0.15627744
## 51 Cranberry Strawberry dashed -0.16810249 0.1398057 0.72513486
## 67 Strawberry Cranberry dashed -0.25158762 0.1712755 0.15709032
## 72 Strawberry Cranberry dashed 0.04303751 0.1441119 0.56023100
## 73 Strawberry Cranberry dashed 0.52109773 0.1254820 0.36640738
## sd_symp N_symp N_sumsympallop
## 39 0.1193552 30 60
## 42 0.1531197 30 60
## 43 0.1156381 30 60
## 46 0.1156695 30 60
## 51 0.1275430 30 60
## 67 0.1411560 30 60
## 72 0.1284283 30 60
## 73 0.1201170 30 60TEMP_dataG29_StrawChe## Line Treatment Fruit_s Generation N_allop logchange sd_logchange SA
## 14 CEA Strawberry Cherry 29 30 0.01971359 0.1475176 0
## 16 CEB Strawberry Cherry 29 30 -0.15850680 0.0983504 0
## 19 CEC Strawberry Cherry 29 30 0.21131464 0.1237821 0
## 68 FRA Cherry Strawberry 29 30 -0.50599095 0.1378652 0
## 71 FRB Cherry Strawberry 29 30 0.19377147 0.1094819 0
## 75 FRC Cherry Strawberry 29 30 0.32151693 0.1140574 0
## Symp Allop Line_type logchange_allop sd_allop logchange_symp
## 14 Cherry Strawberry dashed 0.01971359 0.1475176 0.1476700
## 16 Cherry Strawberry dashed -0.15850680 0.0983504 0.2503803
## 19 Cherry Strawberry dashed 0.21131464 0.1237821 0.3585406
## 68 Strawberry Cherry dashed -0.50599095 0.1378652 0.1570903
## 71 Strawberry Cherry dashed 0.19377147 0.1094819 0.5602310
## 75 Strawberry Cherry dashed 0.32151693 0.1140574 0.3664074
## sd_symp N_symp N_sumsympallop
## 14 0.1574284 30 60
## 16 0.1009864 30 60
## 19 0.1169300 30 60
## 68 0.1411560 30 60
## 71 0.1284283 30 60
## 75 0.1201170 30 60# Limits
ymin_CheCranG29=min(min(TEMP_dataG29_CheCran$logchange_allop-1.96*TEMP_dataG29_CheCran$sd_allop, na.rm= TRUE),
min(TEMP_dataG29_CheCran$logchange_symp-1.96*TEMP_dataG29_CheCran$sd_symp, na.rm= TRUE))
ymax_CheCranG29=max(max(TEMP_dataG29_CheCran$logchange_allop + 1.96*TEMP_dataG29_CheCran$sd_allop, na.rm= TRUE),
max(TEMP_dataG29_CheCran$logchange_symp + 1.96*TEMP_dataG29_CheCran$sd_symp, na.rm= TRUE))
lim_text<-ymin_CheCranG29+0.99*(ymax_CheCranG29-ymin_CheCranG29)
CheCran_G29 <- ggplot(data = TEMP_dataG29_CheCran) +
geom_abline(intercept = Estimates_pairwise$Estimates[Estimates_pairwise$Generation ==29 &
Estimates_pairwise$Pairwise == "Cherry_Cranberry" &
Estimates_pairwise$Variables == "intercept"],
slope = Estimates_pairwise$Estimates[Estimates_pairwise$Generation ==29 &
Estimates_pairwise$Pairwise == "Cherry_Cranberry" &
Estimates_pairwise$Variables == "slope"],
colour = "black", size = 0.75) +
geom_errorbar(aes(x =logchange_symp, ymin = logchange_allop-(1.96*sd_allop),
ymax = logchange_allop + (1.96*sd_allop),
color = Symp),
width= 0.02, size = 0.3, alpha = 0.8) +
geom_errorbarh(aes(y = logchange_allop, xmin = logchange_symp-1.96*sd_symp, xmax = logchange_symp + 1.96*sd_symp,
color = Symp),
height = 0.02, size = 0.3, alpha = 0.8) +
geom_point(aes(x =logchange_symp, y = logchange_allop, color = Symp,fill = Symp, shape = Allop),
size =3, fill = "white", stroke =1.2) +
geom_text(x = 0.33, y = lim_text, label = eq_r(gen = 29, pair = "Cherry_Cranberry"), parse = TRUE, color="black", size = 3.5) +
xlab("Fitness change in\nselective environment") +
ylab("Fitness change in\nalternative environment") +
ggtitle("Cherry vs. Cranberry") +
labs(shape = "Test fruit", color = "Evolution fruit") +
scale_shape_manual(values = c(16, 15)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424")) +
coord_cartesian(ylim = c(ymin_CheCranG29, ymax_CheCranG29),
xlim = c(ymin_CheCranG29, ymax_CheCranG29)) +
theme_LO_sober
CheCran_G29
# Limits
ymin_CranStrawG29=min(min(TEMP_dataG29_CranStraw$logchange_allop-1.96*TEMP_dataG29_CranStraw$sd_allop, na.rm= TRUE),
min(TEMP_dataG29_CranStraw$logchange_symp-1.96*TEMP_dataG29_CranStraw$sd_symp, na.rm= TRUE))
ymax_CranStrawG29=max(max(TEMP_dataG29_CranStraw$logchange_allop + 1.96*TEMP_dataG29_CranStraw$sd_allop, na.rm= TRUE),
max(TEMP_dataG29_CranStraw$logchange_symp + 1.96*TEMP_dataG29_CranStraw$sd_symp, na.rm= TRUE))
lim_text<-ymin_CranStrawG29+0.99*(ymax_CranStrawG29-ymin_CranStrawG29)
#Plot
CranStraw_G29 <- ggplot(data = TEMP_dataG29_CranStraw) +
geom_abline(intercept = Estimates_pairwise$Estimates[Estimates_pairwise$Generation ==29 &
Estimates_pairwise$Pairwise == "Cranberry_Strawberry" &
Estimates_pairwise$Variables == "intercept"],
slope = Estimates_pairwise$Estimates[Estimates_pairwise$Generation ==29 &
Estimates_pairwise$Pairwise == "Cranberry_Strawberry" &
Estimates_pairwise$Variables == "slope"],
colour = "black", size = 0.75) +
geom_errorbar(aes(x =logchange_symp, ymin = logchange_allop-(1.96*sd_allop),
ymax = logchange_allop + (1.96*sd_allop),
color = Symp),
width= 0.02, size = 0.3, alpha = 0.8) +
geom_errorbarh(aes(y = logchange_allop, xmin = logchange_symp-1.96*sd_symp, xmax = logchange_symp + 1.96*sd_symp,
color = Symp),
height = 0.02, size = 0.3, alpha = 0.8) +
geom_point(aes(x =logchange_symp, y = logchange_allop, color = Symp,fill = Symp, shape = Allop),
size =3, fill = "white", stroke =1.2) +
geom_text(x = -0.45, y = lim_text, label = eq_r(gen = 29, pair = "Cranberry_Strawberry"), parse = TRUE, color="black", size = 3.5) +
xlab("Fitness change in\nselective environment") +
ylab("Fitness change in\nalternative environment") +
ggtitle("Cranberry vs. Strawberry") +
coord_cartesian(ylim = c(ymin_CranStrawG29, ymax_CranStrawG29),
xlim = c(ymin_CranStrawG29, ymax_CranStrawG29)) +
labs(shape = "Test fruit", color = "Selection fruit") +
scale_shape_manual(values = c(15, 17)) +
scale_color_manual(values = c("#FDB424", "#3FAA96")) +
theme_LO_sober
CranStraw_G29
# Limits
ymin_StrawCheG29=min(min(TEMP_dataG29_StrawChe$logchange_allop-1.96*TEMP_dataG29_StrawChe$sd_allop, na.rm= TRUE),
min(TEMP_dataG29_StrawChe$logchange_symp-1.96*TEMP_dataG29_StrawChe$sd_symp, na.rm= TRUE))
ymax_StrawCheG29=max(max(TEMP_dataG29_StrawChe$logchange_allop + 1.96*TEMP_dataG29_StrawChe$sd_allop, na.rm= TRUE),
max(TEMP_dataG29_StrawChe$logchange_symp + 1.96*TEMP_dataG29_StrawChe$sd_symp, na.rm= TRUE))
lim_text<-ymin_StrawCheG29+0.99*(ymax_StrawCheG29-ymin_StrawCheG29)
StrawChe_G29 <- ggplot(data = TEMP_dataG29_StrawChe) +
geom_abline(intercept = Estimates_pairwise$Estimates[Estimates_pairwise$Generation ==29 &
Estimates_pairwise$Pairwise == "Strawberry_Cherry" &
Estimates_pairwise$Variables == "intercept"],
slope = Estimates_pairwise$Estimates[Estimates_pairwise$Generation ==29 &
Estimates_pairwise$Pairwise == "Strawberry_Cherry" &
Estimates_pairwise$Variables == "slope"],
colour = "black", size = 0.75) +
geom_errorbar(aes(x =logchange_symp, ymin = logchange_allop-(1.96*sd_allop),
ymax = logchange_allop + (1.96*sd_allop),
color = Symp),
width= 0.02, size = 0.2, alpha =1) +
geom_errorbarh(aes(y = logchange_allop, xmin = logchange_symp-1.96*sd_symp, xmax = logchange_symp + 1.96*sd_symp,
color = Symp),
height = 0.02, size = 0.2, alpha =1) +
geom_point(aes(x =logchange_symp, y = logchange_allop, color = Symp,fill = Symp, shape = Allop),
size =3, fill = "white", stroke =1.2) +
geom_text(x = -0.2, y = lim_text, label = eq_r(gen = 29, pair = "Strawberry_Cherry"), parse = TRUE, color="black", size = 3.5) +
coord_cartesian(ylim = c(ymin_StrawCheG29, ymax_StrawCheG29),
xlim = c(ymin_StrawCheG29, ymax_StrawCheG29)) +
xlab("Fitness change in\nselective environment") +
ylab("Fitness change in\nalternative environment") +
ggtitle("Strawberry vs. Cherry") +
labs(shape = "Test fruit", color = "Selection fruit") +
scale_shape_manual(values = c(16, 17)) +
scale_color_manual(values = c("#BC3C6D", "#3FAA96")) +
theme_LO_sober
StrawChe_G29
legend_tradevide <- ggplot(data = data_logchange[data_logchange$Generation == "29",],
aes(x =logchange, y = logchange, color = Symp,fill = Symp, shape = Allop)) +
geom_point(size =2.5, fill = "white") +
labs(shape = "Test fruit", color = "Selection fruit") +
scale_shape_manual(values = c(16, 15, 17)) +
scale_color_manual(values = c("#BC3C6D", "#FDB424", "#3FAA96")) +
theme_LO_sober
legend_trade <- lemon::g_legend(legend_tradevide)
Slopeestimates_logchange_pairwise_errorbar <- cowplot::ggdraw() +
cowplot::draw_plot(CheCran_G7 + theme(legend.position = "none"),
x = 0.01, y = 0.5, width = 0.22, height = 0.45) +
cowplot::draw_plot(CranStraw_G7 + theme(legend.position = "none"),
x = 0.31, y = 0.5, width = 0.22, height = 0.45) +
cowplot::draw_plot(StrawChe_G7 + theme(legend.position = "none"),
x = 0.61, y = 0.5, width = 0.22, height = 0.45) +
cowplot::draw_plot(legend_trade, x = 0.85, y = 0.5, width = 0.1, height = 0.1) +
cowplot::draw_plot(CheCran_G29 + theme(legend.position = "none"),
x = 0.01, y = 0, width = 0.22, height = 0.45) +
cowplot::draw_plot(CranStraw_G29 + theme(legend.position = "none"),
x = 0.31, y = 0, width = 0.22, height = 0.45) +
cowplot::draw_plot(StrawChe_G29 + theme(legend.position = "none"),
x = 0.61, y = 0, width = 0.22, height = 0.45) +
cowplot::draw_plot_label(c("Intermediate phenotyping step", "A", "B", "C", " ",
"Final phenotyping step", "D", "E", "F", " "),
x = c(0.30, 0.01, 0.30, 0.61, 0.92, 0.250, 0.01, 0.30, 0.61, 0.92),
y = c(0.99, 0.95, 0.95, 0.95, 0.95, 0.49, 0.45, 0.45, 0.45, 0.45),
hjust = c(-0.25, -0.25, -0.25, -0.25, -0.25, -0.75, -0.75, -0.75, -0.75, -0.75),
vjust = c(1.5, 1.5, 1.5, 1.5, 1.5, 1.5, 1.5, 1.5, 1.5, 1.5),
size = 16)
Slopeestimates_logchange_pairwise_errorbar
#
# #
#
cowplot::save_plot(file =here::here("figures", "FIG_Correlation.pdf"),
Slopeestimates_logchange_pairwise_errorbar,
base_height = 18/cm(1), base_width = 34/cm(1), dpi = 610)6 SA AND LOCAL ADAPTATION
6.1 Analysis fitness change
# Add vector of weight
data_logchange$Inv_sum<-NA
data_logchange$Vector_sample<-NA
data_logchange[data_logchange$Generation=="7",]$Inv_sum <- 1 / (data_logchange[data_logchange$Generation=="7",]$sd_logchange^2)
data_logchange[data_logchange$Generation=="7",]$Vector_sample <- data_logchange[data_logchange$Generation=="7",]$Inv_sum /
sum(data_logchange[data_logchange$Generation=="7",]$Inv_sum,na.rm = TRUE)
data_logchange[data_logchange$Generation=="29",]$Inv_sum <- 1 / (data_logchange[data_logchange$Generation=="29",]$sd_logchange^2)
data_logchange[data_logchange$Generation=="29",]$Vector_sample <- data_logchange[data_logchange$Generation=="29",]$Inv_sum /
sum(data_logchange[data_logchange$Generation=="29",]$Inv_sum,na.rm = TRUE)
# Problem with NA (if one sd missing): replace by 0 (max value)
data_logchange$Vector_sample[is.na(data_logchange$Vector_sample)] <- 0
#####
##### G7
#####
lm_val_G7 = lm(logchange ~ Treatment + Line + SA + Treatment:Fruit_s,
weights = Vector_sample, data = data_logchange[data_logchange$Generation=="7",])
Fratio = anova(lm_val_G7)[3, 3]/anova(lm_val_G7)[4, 3]
pvalue = 1 - pf(Fratio, anova(lm_val_G7)[3, 1], anova(lm_val_G7)[4, 1])
pvalue## [1] 0.9777576Fratio## [1] 0.0009158865anova(lm_val_G7)[3, 1]## [1] 1anova(lm_val_G7)[4, 1]## [1] 3#####
##### G29
#####
lm_val_G29 = lm(logchange ~ Treatment + Line + SA + Treatment:Fruit_s,
weights = Vector_sample, data = data_logchange[data_logchange$Generation=="29",])
Fratio = anova(lm_val_G29)[3, 3]/anova(lm_val_G29)[4, 3]
pvalue = 1 - pf(Fratio, anova(lm_val_G29)[3, 1], anova(lm_val_G29)[4, 1])
pvalue## [1] 0.211062Fratio## [1] 2.513403anova(lm_val_G29)[3, 1]## [1] 1anova(lm_val_G29)[4, 1]## [1] 3##### G29
#####
lm_val_G29 = lm(logchange ~ Treatment ,
weights = Vector_sample, data = data_logchange[data_logchange$Generation=="29"&
data_logchange$SA=="1",])
lm_val_G29_avec = lm(logchange ~ Treatment + Line ,
weights = Vector_sample, data = data_logchange[data_logchange$Generation=="29"&
data_logchange$SA=="1",])
anova(lm_val_G29,lm_val_G29_avec)## Analysis of Variance Table
##
## Model 1: logchange ~ Treatment
## Model 2: logchange ~ Treatment + Line
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 8 0.018934
## 2 0 0.000000 8 0.018934summary(lm_val_G29_avec)##
## Call:
## lm(formula = logchange ~ Treatment + Line, data = data_logchange[data_logchange$Generation ==
## "29" & data_logchange$SA == "1", ], weights = Vector_sample)
##
## Weighted Residuals:
## ALL 11 residuals are 0: no residual degrees of freedom!
##
## Coefficients: (2 not defined because of singularities)
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.1477 NA NA NA
## TreatmentCranberry 0.5775 NA NA NA
## TreatmentStrawberry 0.2187 NA NA NA
## LineCEB 0.1027 NA NA NA
## LineCEC 0.2109 NA NA NA
## LineCRA -0.8740 NA NA NA
## LineCRB -0.4279 NA NA NA
## LineCRC -0.6981 NA NA NA
## LineCRD -0.8814 NA NA NA
## LineCRE NA NA NA NA
## LineFRA -0.2093 NA NA NA
## LineFRB 0.1938 NA NA NA
## LineFRC NA NA NA NA
##
## Residual standard error: NaN on 0 degrees of freedom
## Multiple R-squared: 1, Adjusted R-squared: NaN
## F-statistic: NaN on 10 and 0 DF, p-value: NAFratio = anova(lm_val_G29)[3, 3]/anova(lm_val_G29)[4, 3]
pvalue = 1 - pf(Fratio, anova(lm_val_G29)[3, 1], anova(lm_val_G29)[4, 1])
pvalue## [1] NAFratio## [1] NAanova(lm_val_G29)[3, 1]## [1] NAanova(lm_val_G29)[4, 1]## [1] NA